Upload
others
View
13
Download
0
Embed Size (px)
Citation preview
GENERALIZED PUSHOVER ANALYSIS FOR UNSYMMETRICAL-PLAN BUILDINGS
A THESIS SUBMITTED TOTHE GRADUATE SCHOOL OF NATURAL AND APPLIED SCIENCES
OFMIDDLE EAST TECHNICAL UNIVERSITY
BY
KAAN KAATSIZ
IN PARTIAL FULFILLMENT OF THE REQUIREMENTSFOR
THE DEGREE OF MASTER OF SCIENCEIN
CIVIL ENGINEERING
JULY 2012
Approval of the thesis:
GENERALIZED PUSHOVER ANALYSIS FOR UNSYMMETRICAL-PLAN BUILDINGS
submitted by KAAN KAATSIZ in partial fulfillment of the requirements for the degree ofMaster of Science in Civil Engineering Department, Middle East Technical Universityby,
Prof. Dr. Canan ÖzgenDean, Graduate School of Natural and Applied Sciences
Prof. Dr. Güney ÖzcebeHead of Department, Civil Engineering
Prof. Dr. Haluk SucuogluSupervisor, Civil Engineering Department, METU
Examining Committee Members:
Prof. Dr. Polat GülkanCivil Engineering Dept., Çankaya University
Prof. Dr. Haluk SucuogluCivil Engineering Dept., METU
Assoc. Prof. Dr. Murat Altug ErberikCivil Engineering Dept., METU
Assoc. Prof. Dr. Afsin SarıtasCivil Engineering Dept., METU
Joseph Kubin, M.Sc.Civil Engineer, PROTA
Date: 17.07.2012
I hereby declare that all information in this document has been obtained and presentedin accordance with academic rules and ethical conduct. I also declare that, as requiredby these rules and conduct, I have fully cited and referenced all material and results thatare not original to this work.
Name, Last Name: KAAN KAATSIZ
Signature :
iii
ABSTRACT
GENERALIZED PUSHOVER ANALYSIS FOR UNSYMMETRICAL-PLAN BUILDINGS
Kaatsız, Kaan
M.S., Department of Civil Engineering
Supervisor : Prof. Dr. Haluk Sucuoglu
July 2012, 104 pages
Nonlinear response history analysis is regarded as the most accurate analysis procedure for
estimating seismic response. Approximate analysis procedures are also available for the deter-
mination of seismic response and they are preferred over nonlinear response history analysis
since much less computational effort is required and good response prediction is achieved by
employing rather simple concepts.
A generalized pushover analysis procedure is developed in this thesis study as an approxi-
mate analysis tool for estimating the inelastic seismic response of structures under earthquake
ground excitations. The procedure consists of applying generalized force vectors to the st-
ructure in an incremental form until a prescribed target interstory drift demand is achieved.
Corresponding generalized force vectors are derived according to this target drift parameter
and include the contribution of all modes. Unlike many approximate analysis procedures, res-
ponse of the structure is directly obtained from generalized pushover analysis results without
employing a modal combination rule, eliminating the errors cultivating from these methods.
Compared to nonlinear response history analysis, generalized pushover analysis is less de-
manding in computational effort and its implementation is simpler relative to other approxi-
mate analysis procedures. It is observed that the proposed analysis procedure yields results
iv
accurately in comparison to the other nonlinear pushover analysis methods. Accordingly it
can be suggested as a convenient and sound analysis tool.
Keywords: Structural evaluation, multi modal pushover analysis, generalized force vectors
v
ÖZ
ÜÇ BOYUTLU BURULMALI SISTEMLER IÇIN GENEL ITME ANALIZI
Kaatsız, Kaan
Yüksek Lisans, Insaat Mühendisligi Bölümü
Tez Yöneticisi : Prof. Dr. Haluk Sucuoglu
Temmuz 2012, 104 sayfa
Yapıların sismik tepkilerin belirlenmesinde, dogrusal olmayan dinamik analiz en kesin sonuç
veren analiz yöntemi olarak kabul görmektedir. Bu yöntemin yanında, sismik tepkilerin tah-
min edilebilmesi için yaklasık analiz yöntemleri de mevcuttur. Yaklasık analiz yöntemleri ile
nispeten basit kavramlar kullanılarak sismik yapı davranısı yeterli dogrulukta kestirilebilmek-
tedir. Ayrıca bu yöntemler çok daha az islem yükü gerektirdigi ve uygulama basitligine sahip
oldukları için pratikte dogrusal olmayan dinamik analiz yöntemi yerine tercih edilmektedirler.
Bu tez çalısmasında yapıların deprem yer hareketleri altında dogrusal olmayan sismik tepkile-
rini tahmin edebilmek için genel itme analizi yöntemi gelistirilmistir. Yaklasık bir analiz yön-
temi olan genel itme analizi, dogrusal olmayan statik itme analizi prensibini kullanmaktadır.
Bu yöntemde, mod birlestirme kuralları kullanarak hesaplanan genel kuvvet vektörleri yapıya
artımsal olarak bir katın önceden belirlenmis kat arası öteleme talebine ulasıncaya kadar uy-
gulanmaktadır. Bir kat ile ilgili olarak hesaplanan genel yük vektörleri, bir kat arası öteleme
parametresine göre türetilir ve tüm modların katkısını içerir. Pek çok yaklasık analiz yöntemi-
nin aksine, yapının tepkisi direk olarak genel itme analizi sonuçlarından elde edilmektedir ve
ilaveten herhangi bir mod birlestirme kuralının uygulanmasına ihtiyaç duymamaktadır. Bu sa-
yede, dogrusal olmayan tepki parametrelerinin mod birlestirme kuralı ile birlestirilmesinden
vi
kaynaklanan hatalar ortadan kaldırılmaktadır. Dogrusal olmayan dinamik analiz ile karsılastı-
rıldıgında, genel itme analizinin islemsel güç gereksinimi çok daha azdır ve diger çok modlu
statik itme analizi yöntemlerine göre uygulanması daha basittir. Önerilen analiz yöntemi diger
çok modlu dogrusal olmayan itme analizi yöntemleri ile kıyaslandıgında daha dogru sonuç-
lar üretmektedir. Bu sebeplerle, genel itme analizi kullanıslı ve güvenilir bir analiz yöntemi
olarak önerilmektedir.
Anahtar Kelimeler: Yapısal degerlendirme, çok modlu statik itme analizi, genel yük vektörleri
vii
To my mother...
viii
ACKNOWLEDGMENTS
This study was conducted under the supervision of Prof. Dr. Haluk Sucuoglu. I would like
to express my earnest thanks and appreciations for his support, guidance, encouragement and
criticisms during this study. It was a great honor and pleasure to work with him.
I would like to thank to my family for their endless support and love. During my study,
I felt their encouragement and guidance all the time. Their enthusiasm about this study was
always my main source of motivation. The patience and support shown by them are thankfully
acknowledged.
I also want to extend my thanks to my office mates M. Basar Mutlu, F. Soner Alıcı, M. Can
Yücel, Ahmet Kusyılmaz, Sadun Tanıser and Alper Ö. Gür. The joyful times that we shared
in Room Z01 and their enormous support will always be remembered with pleasure.
Sincere thanks to M. Basar Mutlu and F. Soner Alıcı for their great friendship. Long, tiresome
yet productive and cheerful study sessions we had together over the years, their assistance in
hard times and great memories that we shared together are gladly remembered.
Finally, my very special thanks go to my mother, who passed away a short time ago before the
completion of this study. Her great love and guidance have been essential for me throughout
my life. She always put me before herself and even during her illness she wanted me to
concentrate on my work. Her self-devotion, love, encouragement, support and guidance are
greatly acknowledged. Even though she deceased at the final phase of the study, I am sure that
she sees what her son achieved and she is proud of me. This study is dedicated to her.
ix
TABLE OF CONTENTS
ABSTRACT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . iv
ÖZ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . vi
ACKNOWLEDGMENTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ix
TABLE OF CONTENTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . x
LIST OF TABLES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xii
LIST OF FIGURES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xiii
CHAPTERS
1 INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.1 Statement of the Problem . . . . . . . . . . . . . . . . . . . . . . . 1
1.2 Review of Past Studies . . . . . . . . . . . . . . . . . . . . . . . . . 2
1.2.1 Previous Studies on Nonlinear Static Analysis Procedures 2
1.2.2 Previous Studies on Multi Modal Pushover Analysis Pro-cedures . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
1.2.3 Previous Studies on Generalized Force Vectors . . . . . . 7
1.2.4 Previous Studies on Seismic Analysis of Unsymmetrical-Plan Buildings . . . . . . . . . . . . . . . . . . . . . . . 8
1.3 Objective and Scope . . . . . . . . . . . . . . . . . . . . . . . . . . 10
2 GENERALIZED PUSHOVER ANALYSIS FOR TORSIONALLY COUP-LED SYSTEMS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
2.2 Generalized Force Vectors for Torsionally Coupled Systems . . . . 13
2.3 Target Interstory Drift Demand . . . . . . . . . . . . . . . . . . . . 18
2.4 GPA Formulation for Frames in 3D Torsionally Coupled Structures . 18
2.5 Simplified Implementation of GPA in Torsionally Coupled Systems . 24
2.6 Simplified GPA Procedure . . . . . . . . . . . . . . . . . . . . . . 25
x
3 EARTHQUAKE GROUND MOTIONS . . . . . . . . . . . . . . . . . . . . 28
4 CASE STUDY I: EIGHT STORY UNSYMMETRICAL-PLAN BUILDING . 32
4.1 General Information and Modeling of the Building . . . . . . . . . . 32
4.2 Free Vibration Properties . . . . . . . . . . . . . . . . . . . . . . . 36
4.3 Presentations of the Analysis Results . . . . . . . . . . . . . . . . . 36
5 CASE STUDY II: TWENTY STORY UNSYMMETRICAL-PLAN BUIL-DING . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72
5.1 General Information and Modeling of the Building . . . . . . . . . . 72
5.2 Free Vibration Properties . . . . . . . . . . . . . . . . . . . . . . . 77
5.3 Issues Encountered During Analyses . . . . . . . . . . . . . . . . . 78
5.4 Presentation of the Analysis Results . . . . . . . . . . . . . . . . . . 79
6 SUMMARY and CONCLUSIONS . . . . . . . . . . . . . . . . . . . . . . . 97
6.1 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
6.2 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98
REFERENCES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100
xi
LIST OF TABLES
TABLES
Table 2.1 tmax values (seconds) for maximum interstory drifts of an eight story tor-
sionally coupled structure composed of four frames in the direction of analysis
(Figure 2.2). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
Table 3.1 Selected ground motions and their properties . . . . . . . . . . . . . . . . 29
Table 4.1 Free vibration properties of the eight story structure . . . . . . . . . . . . . 37
Table 5.1 Free vibration properties of the twenty story structure . . . . . . . . . . . . 78
xii
LIST OF FIGURES
FIGURES
Figure 2.1 Effective force vector for second story displacement . . . . . . . . . . . . 14
Figure 2.2 Plan view of the eight story unsymmetrical-plan structure (all units in meters). 19
Figure 2.3 Maximum interstory drifts of the 8-Story structure determined from LRHA
(plan view) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
Figure 2.4 Location of Frame k with respect to the center of mass . . . . . . . . . . . 21
Figure 3.1 Acceleration-time histories of the selected ground motions . . . . . . . . . 30
Figure 3.2 Acceleration response spectra of the selected ground motions . . . . . . . 31
Figure 4.1 Elevation view of the frames in the direction of analysis (all units in meters). 33
Figure 4.2 Interstory drift ratios of four frames under GM1 . . . . . . . . . . . . . . 38
Figure 4.3 Mean beam plastic rotations of four frames under GM1 . . . . . . . . . . 38
Figure 4.4 2’nd and 6’th story beam plastic rotations of four frames under GM1 . . . 39
Figure 4.5 Mean column chord rotations of four frames under GM1 . . . . . . . . . . 40
Figure 4.6 2’nd and 6’th story beam end moments of four frames under GM1 . . . . . 41
Figure 4.7 Story shear forces of four frames under GM1 . . . . . . . . . . . . . . . . 42
Figure 4.8 Interstory drift ratios of four frames under GM2 . . . . . . . . . . . . . . 43
Figure 4.9 Mean beam plastic rotations of four frames under GM2 . . . . . . . . . . 43
Figure 4.10 2’nd and 6’th story beam plastic rotations of four frames under GM2 . . . 44
Figure 4.11 Mean column chord rotations of four frames under GM2 . . . . . . . . . . 45
Figure 4.12 2’nd and 6’th story beam end moments of four frames under GM2 . . . . . 46
Figure 4.13 Story shear forces of four frames under GM2 . . . . . . . . . . . . . . . . 47
Figure 4.14 Interstory drift ratios of four frames under GM3 . . . . . . . . . . . . . . 48
Figure 4.15 Mean beam plastic rotations of four frames under GM3 . . . . . . . . . . 48
xiii
Figure 4.16 Mean column chord rotations of four frames under GM3 . . . . . . . . . . 49
Figure 4.17 2’nd and 6’th story beam end moments of four frames under GM3 . . . . . 50
Figure 4.18 Story shear forces of four frames under GM3 . . . . . . . . . . . . . . . . 51
Figure 4.19 Interstory drift ratios of four frames under GM4 . . . . . . . . . . . . . . 52
Figure 4.20 Mean beam plastic rotations of four frames under GM4 . . . . . . . . . . 52
Figure 4.21 Mean column chord rotations of four frames under GM4 . . . . . . . . . . 53
Figure 4.22 2’nd and 6’th story beam end moments of four frames under GM4 . . . . . 54
Figure 4.23 Story shear forces of four frames under GM4 . . . . . . . . . . . . . . . . 55
Figure 4.24 Interstory drift ratios of four frames under GM5 . . . . . . . . . . . . . . 56
Figure 4.25 Mean beam plastic rotations of four frames under GM5 . . . . . . . . . . 56
Figure 4.26 Mean column chord rotations of four frames under GM5 . . . . . . . . . . 57
Figure 4.27 2’nd and 6’th story beam end moments of four frames under GM5 . . . . . 58
Figure 4.28 Story shear forces of four frames under GM5 . . . . . . . . . . . . . . . . 59
Figure 4.29 Interstory drift ratios of four frames under GM6 . . . . . . . . . . . . . . 60
Figure 4.30 Mean beam plastic rotations of four frames under GM6 . . . . . . . . . . 60
Figure 4.31 Mean column chord rotations of four frames under GM6 . . . . . . . . . . 61
Figure 4.32 2’nd and 6’th story beam end moments of four frames under GM6 . . . . . 62
Figure 4.33 Story shear forces of four frames under GM6 . . . . . . . . . . . . . . . . 63
Figure 4.34 Interstory drift ratios of four frames under GM7 . . . . . . . . . . . . . . 64
Figure 4.35 Mean beam plastic rotations of four frames under GM7 . . . . . . . . . . 64
Figure 4.36 Mean column chord rotations of four frames under GM7 . . . . . . . . . . 65
Figure 4.37 2’nd and 6’th story beam end moments of four frames under GM7 . . . . . 66
Figure 4.38 Story shear forces of four frames under GM7 . . . . . . . . . . . . . . . . 67
Figure 4.39 Interstory drift ratios of four frames under GM8 . . . . . . . . . . . . . . 68
Figure 4.40 Mean beam plastic rotations of four frames under GM8 . . . . . . . . . . 68
Figure 4.41 Mean column chord rotations of four frames under GM8 . . . . . . . . . . 69
Figure 4.42 2’nd and 6’th story beam end moments of four frames under GM8 . . . . . 70
Figure 4.43 Story shear forces of four frames under GM8 . . . . . . . . . . . . . . . . 71
xiv
Figure 5.1 Plan view of the twenty-story unsymmetrical-plan structure (all units in
meters). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
Figure 5.2 Elevation view of the frames in the direction of analysis (all units in meters). 74
Figure 5.3 Shear wall modeling approach employed in the case study. . . . . . . . . . 76
Figure 5.4 Interstory drift ratios of five frames under GM1 . . . . . . . . . . . . . . . 81
Figure 5.5 Mean beam plastic rotations of five frames under GM1 . . . . . . . . . . . 82
Figure 5.6 Mean column chord rotations of five frames under GM1 . . . . . . . . . . 83
Figure 5.7 Y-axis shear wall chord rotations under GM1 . . . . . . . . . . . . . . . . 84
Figure 5.8 Interstory drift ratios of five frames under GM2 . . . . . . . . . . . . . . . 85
Figure 5.9 Mean beam plastic rotations of five frames under GM2 . . . . . . . . . . . 86
Figure 5.10 Mean column chord rotations of five frames under GM2 . . . . . . . . . . 87
Figure 5.11 Y-axis shear wall chord rotations under GM2 . . . . . . . . . . . . . . . . 88
Figure 5.12 Interstory drift ratios of five frames under GM4 . . . . . . . . . . . . . . . 89
Figure 5.13 Mean beam plastic rotations of five frames under GM4 . . . . . . . . . . . 90
Figure 5.14 Mean column chord rotations of five frames under GM4 . . . . . . . . . . 91
Figure 5.15 Y-axis shear wall chord rotations under GM4 . . . . . . . . . . . . . . . . 92
Figure 5.16 Interstory drift ratios of five frames under GM5 . . . . . . . . . . . . . . . 93
Figure 5.17 Mean beam plastic rotations of five frames under GM5 . . . . . . . . . . . 94
Figure 5.18 Mean column chord rotations of five frames under GM5 . . . . . . . . . . 95
Figure 5.19 Y-axis shear wall chord rotations under GM5 . . . . . . . . . . . . . . . . 96
xv
CHAPTER 1
INTRODUCTION
1.1 Statement of the Problem
There are various analysis procedures for the estimation of seismic demands of structures
available in modern engineering practice. They can be mainly grouped as dynamic and static
analysis procedures. Both dynamic and static analysis procedures are further grouped as
linear and nonlinear analysis methods. These procedures vary in performance and computa-
tional effort. Nonlinear dynamic (nonlinear response history) analysis procedure is generally
accepted as the most accurate among the aforementioned analysis methods for predicting
seismic structural response. Although it is accurate, there are several drawbacks which ac-
company this high accuracy. First, its implementation is much more complex than the other
procedures. Second, convergence and stability problems may occur frequently which prevents
achieving correct estimations. Finally, the required amount of computational effort in terms
of run time, computing power and post processing complexity is quite high. Due to these
drawbacks, approximate analysis procedures that yield sufficiently accurate results are usually
preferred in engineering practice. Many of these approximate analysis procedures have also
several shortcomings. For instance, linear and nonlinear static analysis methods mainly em-
ploy statistical modal combination rules in the estimation of actual dynamic response. These
combination rules introduce errors in the obtained results since directional combination of
the modal responses during the actual dynamic response is ignored in statistical combination.
Errors are more prominent in the estimation of member forces since quadratic combination of
modal contributions produces results that are higher than the actual capacities of these mem-
bers. Moreover, many of the proposed nonlinear static analysis procedures are adaptive, that
is, they require an eigenvalue analysis at each static loading increment. As a result, these
1
procedures require developing special computer codes in their implementation. Therefore, a
non-adaptive nonlinear static analysis procedure that includes modal contribution of all modes
without employing statistical combination, and that is conceptually simple and easy to imple-
ment by available structural analysis software may serve as a convenient approach for seismic
demand estimation.
1.2 Review of Past Studies
Generalized pushover analysis procedure that is proposed in this thesis study is a nonlinear
static analysis method. Hence, review of past studies will focus on such analysis procedures.
The review is presented in three sections. In the first section, studies on nonlinear static anal-
ysis procedures are presented. In the second section, studies on more specialized forms of
nonlinear static analysis, namely multi modal pushover analysis procedures are presented.
Third section of the review presents brief information about the past studies completed on the
generalized force vectors, which is a starting point for this thesis study. Since performance
of generalized pushover analysis is evaluated on an unsymmetrical-plan building in this the-
sis study, final section of the review is dedicated to the past studies on seismic analysis of
unsymmetrical plan, or torsionally irregular buildings.
1.2.1 Previous Studies on Nonlinear Static Analysis Procedures
The idea of conducting a nonlinear pushover analysis in order to determine the force defor-
mation characteristics of an equivalent single degree of freedom (SDOF) model was first in-
troduced by Saiidi and Sözen (1981). In the Q-model they have proposed, force-deformation
relationship of the SDOF model is determined by recording the top story displacement and
corresponding overturning moment of a system under a triangular-shaped increasing mono-
tonic loading. Q-model served as a basis for the N2 method proposed by Fajfar and Fischinger
(1988). In the first stage of the method, capacity curve of the multi degree of freedom (MDOF)
system is determined by nonlinear static analysis under a monotonically increasing lateral
force vector. Then, the equivalent SDOF model representing the MDOF system is constructed
by using the capacity curve determined. In the next stage, displacement demand of the SDOF
model is calculated either by employing nonlinear response history analysis of the SDOF sys-
2
tem, or by using an inelastic displacement spectrum. In the final stage, calculated inelastic
demand is converted to the top story displacement of the MDOF system and entire structural
response is obtained from the pushover analysis at the calculated top story displacement. In
order to adapt the lateral force vector to the nonlinear behavior of the MDOF system, Fajfar
and Gaspersic (1996) proposed a force vector proportional to the multiplication of an assumed
displacement shape, namely a post yield displacement shape, with the mass matrix.
Implementation of inelastic demand estimation by using the nonlinear static analysis proce-
dure as a seismic assessment method is also implemented in seismic codes (FEMA-356, 2000;
ATC-40, 1996) . In order to improve the procedure, different methods were proposed for the
determination of displacement demand in a SDOF system. Capacity Spectrum Method pro-
posed by Freeman et al. (1975) employs an equivalent linearization procedure for SDOF de-
mand determination; while displacement coefficient method described in (FEMA-356, 2000)
is based on the idea of multiplying the demand of an elastic SDOF system with conversion
coefficients to calculate the inelastic displacement demand. Considering the advances in the
SDOF demand estimation, Fajfar (2000) proposed a further improvement to N2 Method where
the displacement demands of SDOF systems is determined using the inelastic acceleration-
displacement spectra.
As a seismic performance evaluation method, nonlinear static analysis is discussed in detail by
Krawinkler and Seneviranta (1998). They suggested that unless higher modes do not severely
affect the structure, nonlinear pushover analysis provides a great improvement over elastic
evaluation tools and it can expose design weaknesses of the structures that are not observed
while elastic analysis is employed. Furthermore, estimation of global and local inelastic de-
mands is proved to be sufficient. For inclusion of the effects of the higher modes in the lateral
load vector, studies have also been conducted. Yang and Wang (2000) compared various lat-
eral load distributions and commented that seismic response evaluation with a force vector
constructed by combining the modal forces with square root of the sum of the squares (SRSS)
approach yields better results. Performance of nonlinear static analysis was further evaluated
by Mwafy and Elnashnai (2001) on several reinforced concrete buildings. They stated that
pushover analysis with non-adaptive lateral force distribution produced appropriate results for
low rise and short period buildings and the discrepancies between static and dynamic analysis
results for special and long period buildings are mainly due to the limited capability of the
fixed load distribution in predicting higher mode response in the post-elastic range.
3
As the inability of conventional nonlinear static analysis for capturing the higher mode effects
became apparent with more studies emerging, adaptive pushover analysis procedures were
developed in order to improve the performance of the pushover analysis procedure. An adap-
tive pushover algorithm for an improvement of the capacity spectrum method was proposed
by Bracci et al. (1997). In the algorithm proposed, response of the structure is divided into
four phases, namely, initial elastic state, transition from elastic to inelastic state, formation
of mechanism and failure state. For each phase, separate eigenvalue analysis is conducted
and modal properties are determined, and drift demand curves are constructed using modal
superposition. Gupta and Kunnath (2000) suggested an adaptive spectra-based pushover pro-
cedure. They proposed an algorithm that involves eigenvalue analysis and determination of
modal properties at each step of the lateral load increment. Along with the update of the
modal properties of the structure, spectral demands for each mode are also calculated again
according to this new modal behavior. Modal force vectors with new spectral demands are
constructed for each mode, and nonlinear static analysis is performed independently. Using
the SRSS rule, results of each static analysis is combined and deformations of the structure
for that load increment are obtained. Another adaptive pushover analysis procedure was de-
veloped by Antoniou and Pinho (2004a). They stated that while using combination rules,
equilibrium is not satisfied at each step. Therefore, instead of combining the nonlinear static
analysis results for each mode at the end of each load increment, they proposed combination
of the modal static forces prior to nonlinear static analysis at each step. They stated that a
minor improvement over conventional pushover analysis was achieved. In their companion
paper, Antoniou and Pinho (2004b) developed a new displacement-based adaptive pushover
analysis procedure. A displacement including the contributions from all modes is constructed
through employment of a combination rule and acted upon the structure. They concluded that,
demand estimation improved considerably in comparison to conventional pushover analysis.
Rofooeia et al. (2007) proposed another adaptive pushover analysis algorithm where a spe-
cific load pattern is updated using first few modes. In the formulation of the algorithm, this
update is directly connected to the relative displacement of the roof with respect to the ground.
They concluded that the method has higher accuracy than conventional pushover analysis for
high-rise buildings; however, they state that even this improvement to the pushover analysis
cannot match the results of time history analysis for a high-rise structure.
Nonlinear static analysis procedures become increasingly complex and hard to implement as
4
improvements made. Adaptive pushover analysis is an important enhancement over conven-
tional pushover analysis; however necessity for continuous update of modal properties results
in an increased computational effort. In addition, adaptive or not, conventional nonlinear static
analysis procedures are not all very successful in estimating the effects of higher modes. Ac-
cordingly, new multi modal pushover analysis procedures have been developed over the last
decade.
1.2.2 Previous Studies on Multi Modal Pushover Analysis Procedures
Multi modal pushover analysis is based upon the idea of conducting more than one pushover
analyses with different lateral force distributions which are proportional to the multiplication
of the mass matrix and the mode shape of each different mode. Paret et al. (1996) first sug-
gested this procedure and they stated that after conducting these analyses, capacity curves are
converted to ADRS format (Mahaney 1993). Then, yielding pseudo acceleration values for
the pushover curve of each mode are determined. From the elastic spectrum, elastic pseudo
acceleration values corresponding to the period of each mode are also identified. After these
values are obtained, the Modal Criticality Index (MCI) is calculated for each mode by means
of dividing the elastic pseudo acceleration by yielding pseudo acceleration. The mode which
has the highest MCI value is determined as the critical mode. They concluded that code design
methodologies that are mainly based on conventional pushover analysis do not adequately ad-
dress the higher mode induced story collapses in long-period structures. Sasaki et al. (1998)
later based the Multi-Mode Pushover procedure on the previous MCI determination method.
Using the ADRS format, they obtained the capacity curve of each mode and studied the effect
of higher modes on 17-story and 12-story steel frames. For the 17-story frame, they identified
a high inelastic demand that causes beam hinging in the upper stories. For the 12-story frame,
they also observed significant inelastic demand and accompanying damage. It is pointed out
that for both frames, pushover analysis with triangular load pattern were not successful in the
estimation of these damages.
Modal Pushover Analysis (MPA) is proposed by Chopra and Goel (2002). In the MPA pro-
cedure, pushover analyses are conducted for each mode and modal demands are determined
from response history analyses of inelastic SDOF systems whose capacity curves are con-
structed from the previously performed pushover analyses. Finally, response parameters of the
5
structure at these modal demands are combined by employing SRSS or Complete Quadratic
Combination (CQC) to get the structural response. MPA procedure is based on the assumption
of independent inelastic response for each mode which may not be the case for structures such
as unsymmetrical-plan buildings where coupling of modes is significant. Chintanapakdee and
Chopra (2003) evaluated MPA using regular generic frames and concluded that when a suf-
ficient number of modes are included, drifts determined by MPA is generally similar to the
results from nonlinear response history analysis, while first mode alone pushover procedures
does not display the same accuracy. Also they stated that the deviation from actual response
tends to increase for longer-period frames and larger SDOF-system ductility factors, espe-
cially for taller frames where higher mode distribution are more pronounced. Using vertically
irregular frames, Chintanapakdee and Chopra (2004) also assessed the performance of MPA.
Chopra and Goel (2004) proposed the application of MPA to unsymmetrical-plan buildings.
Upon observing the member forces computed by MPA, Goel and Chopra (2005) stated that
they are unrealistic such that they exceed member capacities. Therefore, they modified the
MPA so that it can estimate the member forces correctly. In their proposed modification, if the
computed member force exceeds the member capacity, it is recomputed from the MPA esti-
mate of member deformations using the member force-deformation relationship. In an effort
to merge the concepts of capacity spectrum method, adaptive pushover analysis procedure
suggested by Gupta and Kunnath (2000) and MPA, Kalkan and Kunnath (2006) proposed a
new pushover analysis procedure. Mao et al. (2008) presented an improved MPA procedure in
which they update the lateral force vector for the first mode yielding point. A new pushover
analysis with the updated force vector is conducted for the first mode; whereas for higher
modes no update is performed. Response is obtained by the modal combination of results
collected from all modes. They concluded that better accuracy is achieved with the improved
MPA procedure. Reyes and Chopra (2011) proposed the three dimensional MPA, where the
analysis algorithm is applied for the two horizontal directions and response obtained is later
combined by the SRSS multi-component combination rule. They also suggested a practical
MPA (PMPA) method which estimates seismic demands directly from the earthquake re-
sponse spectrum rather than response history analysis of the modal SDOF systems for each
ground motion. They stated that the PMPA procedure for nonlinear systems is almost as
accurate as RSA for linear systems.
N2 Method including the effects of higher modes was presented by Kreslin and Fajfar (2011).
6
Taking up the same approach proposed by Fajfar et al. (2005) for asymmetric buildings, their
procedure employs correction factors which are defined as the ratio between the results ob-
tained by elastic modal analysis of the structure and the results obtained by pushover analysis
completed while performing basic N2 method. These correction factors are later used to
determine the local response quantities. They concluded that this extension to N2 method
improved accuracy of the results substantially. They also pointed out that, compared with
the MPA and modified MPA procedures, the extended N2 procedure yielded slightly larger
estimates. These estimates were generally conservative in comparison with the mean results
obtained by nonlinear response history analysis. Recently, Kreslin and Fajfar (2012) com-
bined their previous work on asymmetric buildings in both plan and elevation and extended
the N2 method for these type of structures. The proposed procedure includes using two sets
of correction factors in the extended N2 method, one for displacements and the other one for
story drifts. Correction factor for displacements is obtained from the ratio of the normalized
roof displacements obtained by elastic modal analysis and roof displacements obtained from
pushover analysis. Same procedure is followed for correction factor for story drifts, this time
ratio of story drifts obtained from elastic modal analysis and pushover analysis are calculated.
They concluded that the proposed extended N2 method mostly yields conservative predictions
of higher mode effects.
Jerez and Mebarki (2011) proposed a pseudo-adaptive uncoupled modal response analysis.
They construct an alternative capacity curve for each mode based on the absorbed energy
during pushover analysis. They also state that approximate modal shapes at each pushover
step after yielding can be calculated. Accompanying to these modal shapes, their equivalent
energy-based displacements and corresponding modal participation factors can also be deter-
mined. In their suggested procedure, modal responses obtained using these modal properties
are superimposed rather than employing a modal combination rule and peak responses are
determined.
1.2.3 Previous Studies on Generalized Force Vectors
Generalized force vectors which represent the force vector acting on the system instanta-
neously at the time when a specific response parameter reaches its maximum value were
described by Sucuoglu and Günay (2011). They proposed that these generalized force vectors
7
have the contribution of all modes and if are applied to a structural system, then the maxi-
mum value of the specific response parameter that occurs during dynamic response could be
produced. Selecting the interstory drift as a specific response parameter, these force vectors
were derived. Analytical tests conducted on reinforced concrete frames produced accurate
results in comparison to nonlinear response history analysis. The idea of using generalized
force vectors laid foundation to Generalized Pushover Analysis (GPA) developed in this thesis
study.
Alıcı et al. (2011), further extended the work on generalized force vectors by showing that
using a mean response spectrum for the calculation of force vectors yields similar results to
the mean of the resulting response parameters under a set of strong ground motion excitations.
Hence they have demonstrated the practicality of this approach.
1.2.4 Previous Studies on Seismic Analysis of Unsymmetrical-Plan Buildings
Studies on seismic analysis of unsymmetrical-plan buildings date back to early works on
asymmetric wall-frame structures. Rutenberg and Heidebrecht (1975) proposed an approxi-
mate method for lateral force analysis of asymmetric wall-frame structures. Another approxi-
mate method was suggested by Reinhorn et al. (1977) for the dynamic analysis of torsionally
coupled tall building structures. Previous studies on asymmetric structures were extensively
reviewed by Rutenberg (1992), Rutenberg et al. (1995) and Rutenberg (1998). In his 1998
report, Rutenberg concluded that the studies conducted so far can only be considered as a be-
ginning in understanding the behavior of these structures. Provided that appropriate loading
patterns and eccentricities are selected, he also stated that pushover analysis is a promising
alternative to the linear equivalent lateral force procedure. It is pointed out in his report that
in order to evaluate the approximate methods on asymmetric structures, the problem of de-
termining the displacement target for pushover analysis has to be addressed. Chandler and
Duan (1997) investigated the performance of asymmetric code-designed buildings. Inelastic
earthquake response of single-story asymmetric buildings was worked on by Stathopoulos
and Anagnostopoulos (2003). In their study, performance of simple structures was examined
with the employment of shear-beam type models and plastic hinge idealization of one-story
buildings. Among their conclusions, they deduced that for torsionally flexible unsymmetrical
systems, increase in eccentricity increases ductility demands at the stiff edge and decreases
8
them at the flexible edge. Torsionally unbalanced and irregular concrete buildings have been
the subject of a study conducted by Kosmopoulos and Fardis (2006). According to the provi-
sions of Eurocode 8 (2005), they have developed a computational capability for the analysis,
evaluation and retrofitting of concrete buildings. They verified this computational capability
and modeling approach by comparing the demand estimations for floor displacements and
member damages to the results of pseudo dynamics tests.
In the nonlinear static analysis procedure suggested by Kilar and Fajfar (1997) asymmetric
structures are modeled by using planar macro-elements. For each planar macro-element, a
simple bilinear or multi-linear base shear and top story displacement relationship is assumed
and pushover analysis is conducted. A procedure where higher modal and torsion induced
three-dimensional effects are considered was developed by Moghadam and Tso (1998). In
their procedure, they determined target displacements for resisting elements by conducting
a linear response spectrum analysis of the building. Later, inelastic planar models of these
resisting elements are prepared and they are pushed to these target displacements. Evaluating
the performance of the procedure, they stated that the proposed method leads to good response
estimates for asymmetrical multi-story buildings. A simplified procedure for seismic analysis
of asymmetric plan buildings was proposed by Wilkinson and Thambiratnam (2001) based
on a modification of the shear beam model. According to their conclusions, this modified
shear beam model provides sufficient accuracy and it is applicable to static, quasi-static and
nonlinear dynamic analysis.
Extension of modal pushover analysis to unsymmetrical buildings by Chopra and Goel (2004)
emerged as an important analysis procedure. As difference from 2D case of the procedure,
modal force vectors are composed of lateral forces and torques. When asymmetry is present
in both orthogonal directions, two pushover curves belonging to the structure exist in these
directions. In this case, the authors suggested that pushover curve in the dominant direc-
tion of the mode is to be utilized. After modal pushover analyses with pre-described force
vectors are completed, seismic response of the unsymmetrical-plan structure is obtained by
employing complete quadratic combination (CQC) rule since coupling between modes results
in increased error when SRSS rule is used. The procedure proposed was tested on a mass ec-
centric system for which three different versions were created. MPA is conducted on these
three torsionally stiff, torsionally flexible and torsionally similarly stiff systems and poor per-
formance was observed for the torsionally similarly stiff system due to the strong coupling of
9
modes. Roof displacements were underestimated due to employment of CQC rule.
An extension of the well-known N2 method to asymmetric buildings was proposed by Fajfar
et al. (2005). In this extension, capacity curves are constructed with the N2 method and
seismic demand are determined. Then, torsional effects are determined by a linear modal
analysis of the structure, independently for excitation in two horizontal directions. Results
for the modal analysis are than combined by using SRSS rule. In the final step, structure is
pushed to the demands estimated by N2 method for two horizontal directions and correction
factors to be applied to the results of pushover analyses are determined. These correction
factors are defined as the ratio between the normalized roof displacements obtained by elastic
modal analysis and by pushover analysis. Barros and Almeida (2011) suggested a new multi
modal load pattern based on the relative modal participation of each mode of vibration for
nonlinear static analysis and tested it on two story symmetrical and asymmetric structures.
They defined the modal participation factor they used for determination of the load pattern as
the contribution of each mode to the global response of the system and formed the multi modal
load pattern by multiplying mode shapes by corresponding modal participation factors and
combining them by summation. At the end of their studies, they concluded that while torsional
modes are also significant on the response of structure, the proposed load pattern created much
more accurate results than the conventional load patterns. They also noted that performance of
the proposed load pattern is satisfactory for asymmetric structures; whilst pushover analysis
with a load pattern proportional to the shape of fundamental mode of vibration overestimates
the response of the system.
Upon inspecting the past studies on unsymmetrical-plan buildings, it can be concluded that
there is still much to be determined for the seismic response of these type of structure. Mainly
the methods employed for analysis are generally extended from procedures initially formu-
lated on 2D frames. Therefore, they have some weaknesses in estimating the response of
asymmetric structures on a complete basis.
1.3 Objective and Scope
A generalized pushover analysis procedure which makes use of generalized force vectors
that includes all modal contributions is presented for inelastic seismic response prediction of
10
asymmetric multi degree of freedom three dimensional systems. The procedure is applied
on an eight story unsymmetrical-plan building where significant coupling of modes exist.
The structure is analyzed with various analysis procedures under 10 ground motions. Results
obtained from generalized pushover analysis are compared with nonlinear response history
results. Also included in the comparisons are response predictions from response spectrum
analysis and modal pushover analysis developed by Chopra and Goel (2004). Several re-
sponse parameters such as interstory drift ratios, chord rotations, plastic rotations and member
forces are compared and performance of the procedure is evaluated.
Main objective of the study is to develop a generalized pushover analysis procedure for in-
elastic seismic response prediction of multi degree of freedom three dimensional systems with
asymmetric plan, and test the accuracy of the proposed procedure in predicting the response
parameters.
This thesis is composed of five main chapters. Brief contents are given as follows.
Chapter 1: Statement of the problem and review of past studies on nonlinear
static analysis procedures, multi modal pushover analysis procedures, gener-
alized force vectors and seismic analysis of unsymmetrical plan buildings.
Chapter 2: Detailed explanation and formulation of generalized pushover
analysis procedure.
Chapter 3: Information about ground motions used in analyses.
Chapter 4: Case study of an eight story reinforced concrete unsymmetrical-
plan building with mass asymmetry. Implementation of generalized pushover
analysis and comparison of its results with those of the other seismic demand
determination procedures.
11
Chapter 5: Case study of a twenty story reinforced concrete unsymmetrical-
plan building. Structural system is composed of both moment frames and a
shear wall core offset from center of stiffness. Implementation of generalized
pushover analysis and comparison of its results.
Chapter 6: Summary and conclusions.
12
CHAPTER 2
GENERALIZED PUSHOVER ANALYSIS FOR TORSIONALLY
COUPLED SYSTEMS
2.1 Introduction
Generalized pushover analysis (GPA) for plane frame structures was developed by Sucuoglu
and Günay (2011). GPA is extended to space frames with torsional coupling in this study.
Analytical formulation of generalized pushover analysis for space frames with unsymmetrical
distribution of mass and/or lateral stiffness in plan is presented in this chapter. Definition of
generalized force vectors is given in the first section. The target demand parameters which
are selected as the target interstory drifts and their derivation are explained thoroughly in the
second section. The analysis procedure is outlined in the last section.
2.2 Generalized Force Vectors for Torsionally Coupled Systems
Response parameters achieve their maximum values at different time instants during seismic
response. For a specific response parameter that reaches its maximum value at tmax, there
is an effective force vector acting on the system at that instant (Figure 2.1). This effective
force vector includes contributions from all modes; therefore it is a generalized force vector.
Upon defining the generalized force vector corresponding to a response parameter at tmax and
applying it to the structure, the maximum value of this response parameter can be obtained
by performing an equivalent static analysis. If the system is linear elastic, direct application
of the force vector in a single load step is sufficient for solving the response parameters. For
nonlinear static analysis, however, generalized force vector is applied in an incremental form
13
until a specified target response demand is attained. Interstory drift is selected as the target
response parameter in this study since it gives a good representation of seismic performance
and damage state of the structure at any deformation state. Estimation of the target drift
demand will be discussed in the following sections.
üg
u
feff, 2
(tmax
!!
u2 (tmax)
Figure 2.1: Effective force vector for second story displacement
Generalized force vectors are derived for linear elastic MDOF systems through the application
of modal superposition principle. The maximum value of an arbitrary response parameter can
be obtained at tmax while the system is subjected to an earthquake ground excitation ug(t). The
force vector acting on the system at tmax is defined by the superposition of contributions from
all modes:
f (tmax) =∑
n
fn(tmax) (2.1)
Effective force vector at the n’th mode is given in Equation 2.2 at time tmax:
fn(tmax) = Γn mϕn An(tmax) (2.2)
Parameters in Equation 2.2 are defined below.
Γn = Ln/Mn Ln = ϕTn m l Mn = ϕT
n mϕn (2.3)
14
Here ϕn is the n’th mode shape, m is the mass matrix and l is the influence vector. An(tmax) in
Equation 2.2 can be expressed in terms of the modal displacement Dn at tmax during seismic
response:
An(tmax) = ω2n Dn(tmax) (2.4)
ω2n in Equation 2.4 is the n’th mode vibration frequency, and Dn satisfies the equation of
motion at tmax.
Dn(tmax) + 2 ξn ωn Dn(tmax) + ω2n Dn(tmax) = −ug(tmax) (2.5)
Since Dn(tmax) occurs at a specific time during seismic excitation, it is not possible to deter-
mine it directly from Equation 2.5 if tmax is not known. In the proposed procedure, tmax is
defined as the time when interstory drift (∆ j) of the j’th story reaches its maximum value.
∆ j,max = ∆ j(tmax) (2.6)
The modal expansion of ∆ j(tmax) is given in Equation 2.7:
∆ j(tmax) =∑
n
Γn Dn(tmax) (ϕn, j − ϕn, j−1) (2.7)
Where ϕn, j is the j’th element of the mode shape vector belonging to the n’th mode. Dividing
both sides of Equation 2.7 by ∆ j(tmax) results in the normalized form of this equation:
1 =∑
n
ΓnDn(tmax)∆ j(tmax)
(ϕn, j − ϕn, j−1) (2.8)
The right hand side of Equation 2.8 yields the normalized contribution of each mode n to the
maximum interstory drift of the j’th story at tmax.
While determination of ∆ j(tmax) still depends on tmax, its counterpart in Equation 2.6 can be
estimated through response spectrum analysis (RSA) by employing a modal combination rule.
There are several statistical combination rules available in literature; however since this thesis
15
study focuses on torsionally coupled structures, the selected method should represent this
behavior adequately. Among statistical combination rules, complete quadratic combination
(CQC) is known to be superior to the simpler quadratic combination rule (SRSS) in that
aspect (Wilson et al., 1981). Moreover; even though coupling does not develop in a system,
higher mode effects on seismic response can be evaluated better with CQC compared to SRSS.
Considering these advantages, CQC is chosen as the combination rule employed in RSA.
∆ j,max is expressed in terms of the modal spectral responses obtained with RSA, combined
with CQC in Equation 2.9:
(∆ j,max)2 =∑i=1
∑n=1
ρin[Γi Di (ϕi, j − ϕi, j−1)
] [Γn Dn (ϕn, j − ϕn, j−1)
](2.9)
Here, ρin is the correlation coefficient. Indices i and j denote corresponding modes and ranges
from 1 to N where N is the number of modes. Dn (or Di) is the spectral displacement of the
n’th (or i’th) mode and it is readily available from the displacement response spectrum of
earthquake ground excitation. A normalized form of Equation 2.9 can also be derived by
simply dividing both sides with (∆ j,max)2. The resulting Equation 2.10 shows the normalized
contribution of combined response of i’th and n’th modes to ∆ j,max which is the maximum
interstory drift ratio at the j’th story.
1 =∑i=1
∑n=1
ρin
(Γi
Di
∆ j,max(ϕi, j − ϕi, j−1)
) (Γn
Dn
∆ j,max(ϕn, j − ϕn, j−1)
)(2.10)
In Equations 2.8 and 2.10, normalized contributions of individual modes to the maximum
interstory drift at a specified story are defined from dynamic response history and response
spectrum analyses, respectively. Under the assumption of equality stated in Equation 2.6, the
right-hand sides of Equations 2.8 and 2.10 can be equated:∑n
ΓnDn(tmax)∆ j(tmax)
(ϕn, j−ϕn, j−1) =∑i=1
∑n=1
ρin
(Γi
Di
∆ j,max(ϕi, j − ϕi, j−1)
) (Γn
Dn
∆ j,max(ϕn, j − ϕn, j−1)
)(2.11)
Leaving out similar terms on both sides of Equation 2.11 results in a simplified form:
Dn(tmax) =∑i=1
ρinDn
∆ j,max
[Γi Di (ϕi, j − ϕi, j−1)
](2.12)
16
The terms in brackets in Equation 2.12 is ∆ j,i, or the i’th mode contribution to the maximum
interstory drift at the j’th story determined from RSA. Inserting ∆ j,i for the bracket term in
Equation 2.12 yields a further simplified expression for Dn(tmax):
Dn(tmax) = Dn
∑i=1ρin ∆ j,i
∆ j,max
(2.13)
Equation 2.13 describes Dn(tmax) independent of tmax through RSA accompanied with CQC.
This equality is designated as the modal scaling rule, since modal displacement amplitude of
the n’th mode at tmax is obtained by the multiplication of spectral displacement of this mode
by a scale factor. The modal scaling factor is defined as the ratio of modal contribution to any
response parameter to the maximum of this response parameter calculated at the j’th story.
Interstory drift in the equation is the response parameter selected for this derivation.
An(tmax) can also be determined from Equation 2.13 by multiplying both sides with ω2n:
An(tmax) = An
∑i=1ρin ∆ j,i
∆ j,max
(2.14)
An is the pseudo-spectral acceleration of the n’th mode and obtained from An = ω2n Dn, similar
to Equation 2.4. fn(tmax) can be rewritten via Equation 2.14 in a form that is independent of
time. Substituting An(tmax) from 2.14 into Equation 2.1 yields:
f (tmax) =∑
n
Γn mϕn An
∑i=1ρin ∆ j,i
∆ j,max
(2.15)
As previously stated, formulation presented herein is based on interstory drift at the j’th story.
Consequently, f (tmax) in Equation 2.15 is the generalized force vector that acts on the system
when the j’th story interstory drift reaches its maximum value. Therefore, f (tmax) will be
denoted as f j in the remaining part of the formulation.
Summation for all modes in Equation 2.15 over all terms in the parentheses and regrouping
leads to a final form to the generalized force vector:
17
f j =∑
n
(Γn mϕn An)
∆ j,n
∆ j,max+
N∑i=1i,n
(ρin
∆ j,i
∆ j,max
) (2.16)
The generalized force vector that maximizes the j’th story interstory drift is presented in
Equation 2.16. This is, in fact, the GPA force vector which is applied on the structure in an
incremental form until the target interstory drift demand at the j’th story is obtained. Details of
the analysis procedure are presented in the following sections, starting with the determination
of target interstory drift demand.
2.3 Target Interstory Drift Demand
Maximum interstory drift demand ∆ j(tmax) during ground motion excitation was defined as a
function of Dn(tmax) in Equation 2.7. ∆ j(tmax) can also be expressed with the implementation
of modal scaling rule by substituting Dn(tmax) from Equation 2.13 into 2.7:
∆ j(tmax) =∑
n
Γn (ϕn, j − ϕn, j−1) Dn
∑i=1ρin ∆ j,i
∆ j,max
(2.17)
As discussed previously, Equation 2.17 expresses target drift at the j’th story which reaches
its maximum value when the corresponding f j is acting on the system. Accordingly, ∆ j(tmax)
can be designated as ∆ jt where t stands for target. Regrouping the terms in Equation 2.17
yields the definition for target interstory drift demand at the center of mass:
∆ jt =∑
n
Γn Dn (ϕn, j − ϕn, j−1)
∆ j,n
∆ j,max+
N∑i=1i,n
(ρin
∆ j,i
∆ j,max
) (2.18)
2.4 GPA Formulation for Frames in 3D Torsionally Coupled Structures
Generalized force vectors and their accompanying target interstory drift demands are deter-
mined by using the eigenvectors defined at the diaphragm centers of mass of a 3D structural
system. For a planar 2D frame or a symmetric-plan 3D structure; it is expected that when
18
target interstory drift for a story is reached during nonlinear static analysis, seismic response
of the entire story can be estimated accordingly. In a 2D frame, all structural members are in
the same frame including the center of mass, resulting in consistent deformations and member
forces at the target interstory demand. Similar to the behavior of 2D frames, response of 3D
plan-symmetric structures also shows no variation within a story since torsional effects are
not present. Consequently, deformations and forces of all structural members in a story can
be estimated accurately by employing a single demand control mechanism, namely using one
target interstory demand for each story. However, this is not the case for 3D buildings where
some form of torsional coupling exists.
Figure 2.2: Plan view of the eight story unsymmetrical-plan structure (all units in meters).
19
It is known that due to strong coupling of modes and torsional effects in unsymmetrical-plan
buildings, maximum values of deformations and forces at different frames within a story occur
at different instants, i.e., at different tmax values during dynamic response. Table 2.1 presents
tmax values at each story for the maximum interstory drifts of four frames of the eight story
unsymmetrical plan building for which the plan view is given in Figure 2.2. These results
have been obtained from linear response history analysis (LERHA) under a ground motion
excitation (Superstition Hills, B-PTS315). Each of these frames is part of the eight story
structure where strong coupling of modes occurs due to a mass eccentricity of 15% at each
story. It is clearly seen that tmax varies with frames at each story.
Table 2.1: tmax values (seconds) for maximum interstory drifts of an eight story torsionallycoupled structure composed of four frames in the direction of analysis (Figure 2.2).
Story Frame FE Frame FI Frame SI Frame SE1 5,01 4,98 4,96 4,842 5,06 5,04 5,00 4,963 5,06 5,04 5,03 5,004 5,07 5,06 5,06 5,065 5,16 5,09 5,09 5,096 5,20 5,19 5,14 5,117 5,22 5,19 5,16 5,148 5,24 5,17 5,16 5,15
In addition to this information, Figure 2.3 shows the maximum interstory drifts of frames in
each story that are determined from LERHA under the same ground motion excitation. It
is not possible to estimate this type of behavior displayed in the figure by representing the
response of each story with a simple 2D formulation. In order to reflect this variation in the
individual frame response, generalized force vectors and target interstory drift demands are
defined for the individual frames of a structure at each story instead of employing a single
force vector and target interstory drift demand couple derived at the center of mass. Multi-
ple tmax instances at each story can be “captured” by this approach, yielding better seismic
response estimations.
20
0
0.01
0.02
0.03
0.04
0.05
0.06
0.07
Insterstory
Drift
(m)
1st Story 2nd Story 3rd Story 4th Story
5th Story 6th Story 7th Story 8th Story
Frame SE Frame SI Frame FI Frame FE
Figure 2.3: Maximum interstory drifts of the 8-Story structure determined from LRHA (planview)
Frame m Frame k
jth Story
!"
CM
!"
#!"
$%"EQ
Figure 2.4: Location of Frame k with respect to the center of mass
21
The n’th mode interstory drifts of Frame k in Figure 2.4 in the direction of ground excitation
is related to the center mass displacements via the distance of frame to the center of mass,
xk along with both translational and rotational components of eigenvectors at the j’th story
(Chopra, 2007). Considering this relation, Equation 2.7 is rewritten as:
∆kj(tmax) =
∑n
Γn Dkn(tmax)
[(ϕny, j − ϕny, j−1
)+ xk
(ϕnθ, j − ϕnθ, j−1
)](2.19)
Dkn(tmax) is the spectral displacement which occurs at the tmax instant for Frame k. Normalized
form of this equation with respect to ∆kj(tmax) is displayed in Equation 2.20:
1 =∑
n
ΓnDk
n(tmax)∆k
j(tmax)
[(ϕny, j − ϕny, j−1
)+ xk
(ϕnθ, j − ϕnθ, j−1
)](2.20)
Similarly, the combined modal response, ∆ j,max which is obtained from RSA through CQC is
defined for Frame k that is shown in Figure 2.4:
(∆kj,max)
2 =∑i=1
∑n=1
ρin
[Γi Di
((ϕiy, j − ϕiy, j−1) + xk (ϕiθ, j − ϕiθ, j−1)
)] [Γn Dn
((ϕny, j − ϕny, j−1) + xk (ϕnθ, j − ϕnθ, j−1)
)](2.21)
Equation 2.21 can also be normalized with respect to (∆ j,max)2:
1 =∑i=1
∑n=1
ρin
ΓiDi
(∆kj,max)
((ϕiy, j − ϕiy, j−1) + xk (ϕiθ, j − ϕiθ, j−1)
) ΓnDn
(∆kj,max)
((ϕny, j − ϕny, j−1) + xk (ϕnθ, j − ϕnθ, j−1)
)(2.22)
The relation defined in Equation 2.6 holds true as in the case of 2D formulation. Taking this
into account, right-hand sides of Equations 2.20 and 2.22 can be equated:
∑n
ΓnDk
n(tmax)∆k
j (tmax)
[(ϕny, j − ϕny, j−1
)+ xk
(ϕnθ, j − ϕnθ, j−1
)]=∑
i=1
∑n=1ρin
[Γi
Di(∆k
j,max)
(ϕiy, j − ϕiy, j−1
)+ xk
(ϕiθ, j − ϕiθ, j−1
)] [Γn
Dn(∆k
j,max)
(ϕny, j − ϕny, j−1
)+ xk
(ϕnθ, j − ϕnθ, j−1
)](2.23)
Normalized form of Equation 2.21 with respect to (∆kj,max)2 is inserted into the right hand side
22
of Equation 2.23 where left hand side represents the normalized form of ∆kj(tmax) in terms of
modal contributions. Leaving out similar terms yields Dkn(tmax) derived for Frame k:
Dkn(tmax) =
∑i=1
ρinDn
∆kj,max
[Γi Di
(ϕiy, j − ϕiy, j−1
)+ xk
(ϕiθ, j − ϕiθ, j−1
)](2.24)
The term in brackets on the right hand side is the i’th mode contribution to the j’th story
interstory drift of Frame k, i.e. ∆kj,i. Then,
Dkn(tmax) = Dn
∑i=1ρin ∆k
j,i
∆kj,max
(2.25)
Here, Equation 2.25 defines Dkn(tmax) in terms of spectral displacement and combined modal
responses. Multiplying both sides by ω2n yields Ak
n(tmax), which, in turn, is used to define the
effective force vector acting at Frame k when the j’th story drift is maximum.
Akn(tmax) = An
∑i=1ρin ∆k
j,i
∆kj,max
(2.26)
Similar to the previously completed derivation, the force vector f k(tmax) acting on Frame k at
the instant when the interstory drift at the j’th story reaches its maximum value is written by
summation of modal contributions. This is achieved by considering the relations defined in
Figure 2.4:
f k(tmax) =∑
n
[Γn mk
(ϕny + xk ϕnθ
)Ak
n(tmax)]
(2.27)
In Equation 2.27, mk is the mass of frame under consideration. In order to obtain the final
form of generalized force vector, Akn(tmax) is inserted from Equation 2.26 into Equation 2.27.
Thus, f kj
for any Frame k is defined as:
f kj =
∑n
Γn mk(ϕny + xk ϕnθ
)Ak
n
∆k
j,n
∆kj,max
+
N∑i=1i,n
ρin
∆kj,i
∆kj,max
(2.28)
23
Equation 2.28 represents the generalized force vector acting at Frame k, when the interstory
drift of the j’th story of that frame reaches its maximum value. In comparison to its 2D
equivalent, Equation 2.16, individual maxima of each frame occurring at different time steps
as in Table 2.1 are taken into account, yielding a more accurate estimation.
Aside from the generalized force vector, the expression for target drift at j’th story of Frame
k is also needed. The modal expansion of the j’th interstory drift occurring at tmax in Frame
k, i.e. ∆kj(tmax) was already defined in Equation 2.19. If Dk
n(tmax) is substituted from Equation
2.25 into Equation 2.19, then ∆kj(tmax) can be expressed in terms of modal spectral responses
by employing the modal scaling rule:
∆kj(tmax) =
∑n
Γn[(ϕny, j − ϕny, j−1
)+ xk
(ϕnθ, j − ϕnθ, j−1
)]Dn
∑i=1ρin ∆k
j,i
∆kj,max
(2.29)
This is in fact the target interstory drift value of the j’th story that is reached when f kj is acting
on Frame k. Similar to 2D formulation, it can be labeled as ∆kjt and expressed in open form in
Equation 2.30:
∆kjt =
∑n
Γn Dn[(ϕiy, j − ϕiy, j−1
)+ xk
(ϕiθ, j − ϕiθ, j−1
)] ∆k
j,n
∆kj,max
+
N∑i=1i,n
ρin
∆kj,i
∆kj,max
(2.30)
Equation 2.30 yields the target interstory drift demand of Frame k at the j’th story under
ground motion excitation. These target interstory drifts are used for demand control at the
k’th frame while conducting pushover analyses with f kj . Therefore f k
j and ∆kjt constitute the
generalized force distribution and the target interstory demand pair for the k’th frame in the
system.
2.5 Simplified Implementation of GPA in Torsionally Coupled Systems
The derivation of two main ingredients of the GPA procedure, GPA force vectors and target
interstory drift demands for 3D torsionally coupled structures have been completed. Building
on these concepts, a simplified implementation of the procedure is introduced in this section.
As discussed in the previous sections, GPA is a nonlinear static analysis procedure relying
24
on multiple pushover analyses conducted by using GPA force vectors derived for each story
of all frames in the system. When the procedure is applied in its complete form, interstory
drift demands at each story and each frame are also defined as target response parameters.
Therefore, the corresponding number of pushovers in GPA is equal to the number of stories
multiplied by number of frames. This, however, creates a significant amount of workload.
For instance, while analyzing an 8-story building with 4 frames in the direction of earthquake
excitation, 32 pushover analyses are needed.
The procedure can be simplified by applying the GPA force vectors at the centers of mass
(Equation 2.16) rather than at each frame individually, while tracking the target interstory
drifts of each frame separately (Equation 2.30). Along a story, different frames reach their
corresponding target interstory drifts at different analysis steps. However, the member re-
sponses obtained at these steps are virtually the same as the ones that are achieved when
performing a full, frame by frame analysis. Application of generalized force vectors defined
at the center of mass significantly reduces the computational effort. To illustrate on the previ-
ously stated 8-story, 4-frame structure, only 8 pushover analyses are performed and 32 target
interstory drift ratios are searched from the frame interstory drifts. Due to the simplicity of-
fered by this approach and the ability to produce results with similar accuracy compared to
frame by frame analysis, the GPA procedure is implemented in this simplified form in the
foregoing presentations.
2.6 Simplified GPA Procedure
The basic approach that the method is built over has been explained in detail above. The
procedure is now developed into an algorithm which can be employed quite easily. At each
pushover analysis conducted, GPA force vector derived for the j’th story of the entire 3D
system (Equation 2.16) is acted upon the structure in an incremental form until the center
of mass reaches the target interstory drift demand calculated according to Equation 2.18. At
each force increment, interstory drifts of all frames at the selected story are determined from
the results recorded during pushover analysis. After they are obtained, target interstory drift
value of each frame is searched. In other words, after the pushover analysis for the j’th story
is performed, interstory drifts for Frame k at each pushover step are browsed in order to find
the step where ∆kjt occurs. When the pushover step is determined, all the deformations and
25
internal forces for structural members of Frame k are acquired from this stage of nonlinear
static analysis. These results represent the response of the structural members when the target
interstory drift, ∆kjt is reached under f j. The global response of the entire structure is estimated
by collecting all member deformations and internal forces at every target interstory drift, ∆kjt
from pushover analyses conducted with each f j and selecting the absolute maximum values
through an envelope algorithm.
Force vectors and target interstory drifts are calculated by using the linear elastic properties of
the structure. It is known that at the advanced stages of pushover analysis, deformations of the
structural system may show quite a different pattern compared to linear elastic behavior due
to nonlinearity. As a result of this phenomenon, the j’th story target interstory drift may not
be achieved occasionally. This situation may sometimes be accompanied with convergence
problems and it may not be possible to achieve the target interstory drift values calculated
by employing linear elastic properties of the structure. In order to overcome this handicap, a
filtering check is performed. After conducting the pushover analysis for the j’th story, it is
checked whether each frame has reached its own ∆kjt, or not. Results of the stories that fail
this check are discarded and the deformations and internal forces of the system are determined
from the envelope results of the pushover analysis of the remaining stories.
To improve the estimation of spectral displacement demand, inelastic spectral displacement,
D∗n may be used for the first two coupled modes in calculating target interstory drifts. In or-
der to determine these first pair of inelastic spectral displacements, modal pushover analyses
of the structure for the first coupled modes are needed. Employing D∗n yields much better
estimation of target interstory drift demand and this in turn increases the chances that conver-
gence problems are not encountered. Therefore, in exchange for a small amount of increased
workload, a good improvement is achieved.
The GPA algorithm which is derived and discussed thoroughly in this chapter is summarized
in the following steps:
1. Modal analysis: Periods (Tn), mode shapes (ϕn) and associated modal properties of the
structure is determined.
2. Response spectrum analysis: Spectral accelerations (An) and displacements (Dn) of
each mode are determined from the corresponding linear elastic spectra. Modal pushover
26
analyses for the first pair of coupled modes can also be conducted in order to estimate
D∗n which is used for estimating the inelastic target interstory drifts. Then, ∆ j,n and
∆ j,max for the center of mass, ∆kj,n and ∆k
j,max for each frame k are obtained by using
these spectral quantities.
3. GPA force vectors: GPA force vectors f j are determined according to Equation 2.16.
4. Target interstory drift demands: Center of mass target interstory drift demands (∆ jt)
which are employed in the incremental nonlinear static analysis under f j are calculated
from Equation 2.18. Frame target interstory drift demands (∆kjt) are also determined by
using Equation 2.30.
5. Nonlinear static analysis: f j are acted upon the system in an incremental form until
∆ jt calculated in step 4 is reached. The convergence check discussed previously is then
employed in order to leave out the stories which have unrealistic responses, if any. Then
the interstory drift record of frame k is searched for ∆kjt. The analysis step that yields
∆kjt is used to compile the member deformations and internal forces at the j’th story
of frame k (target analysis step). This is repeated for every frame where convergence
check has been succeeded.
6. Determination of structural responses: Using the target analysis steps obtained in the
5’th stage of the algorithm, deformations and internal forces of all members at each
frame are determined directly from the deformation state of the structure at the target
interstory drift demand. Entire response of the system is then compiled by employing an
envelope algorithm where absolute maxima of these internal forces and deformations
are selected for every structural member without using any modal combination rule.
These results are listed as final response estimation values.
27
CHAPTER 3
EARTHQUAKE GROUND MOTIONS
The ground motion records which are utilized in the analyses conducted for case studies are
presented in this chapter. Both ordinary and pulse type ground motions are included in the
ground motion set in order to evaluate the performance of Generalized Pushover Analysis
thoroughly under different excitation patterns in case studies. Pulse type ground motions
contain a single significant peak ground acceleration (PGA) value compared to the ordinary
ground motion records. This peak usually occurs in a very short time interval and may stress
the structure considerably during this cycle. In order to estimate the intensity of these pulses,
peak ground velocity values (PGV) are used. Higher PGV usually means a stronger pulse
during the excitation. Ordinary type ground motions, however, lack these distinctive peak
pulses and contain more evenly distributed peaks throughout the record.
Response of the structures to these different types of ground motions may show significant
differences and it is important to consider this behavior. For instance, pulse type ground
motions with a high PGV usually impose a significant demand in a very short time and all
maxima of the member responses are observed almost at this time instant. The instant where
peak demand occurs may result with a substantial amount of yielding and follows with a
change in the modal properties of the structure. Accordingly, the higher mode effects that
are normally anticipated may not be observed. On the other hand, ordinary type ground
motions stress the structure more uniformly during excitation. Since the sudden yielding
accompanying the pulse is absent during the application of the ordinary ground motions, a
better estimation of the individual modal effects can be possible in the analysis.
Eight ground motion records that have been selected in this study are presented in Table 3.1.
All of these records have been downloaded from the PEER strong motion database and used
28
without employing a scaling or any other modification to the original data. For practical
reasons, each ground motion record is given a code number, GM1 to GM8, which will be
used in the remaining part of the study. Second column of the table is the record code that is
associated with the ground motion in the PEER Strong Motion Database. The earthquakes that
these records were produced are given along with the moment magnitude of each earthquake
in the third and fourth columns, respectively. PGA, PGV and peak ground displacement
(PGD) values of each record and type of each ground motion is also shown.
Table 3.1: Selected ground motions and their properties
# GMCODE EARTHQUAKE MwPGA PGV PGD
Type(g) (cm/s) (cm)
GM1 H-E04140 Imperial Valley - 1979 6.5 0.485 37.4 20.1 Pulse
GM2 CLS090 Loma Prieta - 1989 7.0 0.479 45.2 11.3 Pulse
GM3 SPV270 Northridge - 1994 6.7 0.753 84.5 18.7 Pulse
GM4 ORR090 Northridge - 1994 6.7 0.568 51.8 9.0 Ordinary
GM5 ORR360 Northridge - 1994 6.7 0.514 52.0 15.3 Ordinary
GM6 B-PTS315 Superstition Hills - 1987 6.6 0.377 43.9 15.3 Ordinary
GM7 IZT090 Kocaeli - 1999 7.4 0.220 29.8 17.1 Ordinary
GM8 STG000 Loma Prieta - 1989 7.0 0.513 41.2 16.2 Ordinary
Plots of these ground motion records are displayed in Figure 3.1. Acceleration response spec-
tra for 5% damping is shown in Figure 3.2. From the linear elastic spectra, it can be observed
that the ground motions impose different demands on structures with different vibration prop-
erties. In the selection, this property of the ground motions is also considered.
29
0.4
0.2
0.0
0.2
0.4
0.6
0 10 20 30 40
Acceleration!(g)
Time!(Seconds)
GM1
0.4
0.2
0.0
0.2
0.4
0.6
0 10 20 30 40
Acceleration!(g)
Time!(Seconds)
GM2
0.8
0.4
0.0
0.4
0.8
0 5 10 15 20 25
Acceleration!(g)
Time!(Seconds)
GM3
0.4
0.2
0.0
0.2
0.4
0.6
0 10 20 30 40
Acceleration!(g)
Time!(Seconds)
GM4
0.4
0.2
0.0
0.2
0.4
0.6
0 10 20 30 40
Acceleration!(g)
Time!(Seconds)
GM5
0.4
0.2
0.0
0.2
0.4
0.6
0 10 20
Acceleration!(g)
Time!(Seconds)
GM6
0.3
0.1
0.1
0.3
0 5 10 15 20 25 30
Acceleration!(g)
Time!(Seconds)
GM7
0.4
0.2
0.0
0.2
0.4
0.6
0 10 20 30 40
Acceleration!(g)
Time!(Seconds)
GM8
Figure 3.1: Acceleration-time histories of the selected ground motions
30
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
1.8
2
0.00 0.50 1.00 1.50 2.00 2.50 3.00
Acc
eler
atio
n (
g)
Period (Seconds)
GM1 GM2 GM3 GM4 GM5 GM6 GM7 GM8
Figure 3.2: Acceleration response spectra of the selected ground motions
In the following two chapters, case studies of eight-story and twenty-story unsymmetrical-
plan structures are presented. Results of the analyses performed by using the ground motions
given in this chapter and the performance evaluation of GPA are inspected in detail.
31
CHAPTER 4
CASE STUDY I: EIGHT STORY UNSYMMETRICAL-PLAN
BUILDING
4.1 General Information and Modeling of the Building
The first case study is completed on an eight story unsymmetrical-plan reinforced concrete
building and it is presented in this chapter. Plan of the building was presented in Figure 2.2,
and the elevation view of the frames in the direction of analysis is shown in Figure 4.1. The
building is designed in compliance with the TS-500 and Turkish Earthquake Code (TEC). It
is located in the seismic zone 1 where the local site class is defined as Z3. As required by the
regulations stated in the adopted codes, capacity design principles is employed in design, with
an enhanced ductility level of R=8. Material properties are selected as C25 for concrete and
S420 for reinforcing steel. Uniform member dimensions are selected in the design of beams
and columns. Beams are 30x55 cm2 and columns are 50x50 cm2 throughout the building.
Slab thickness at all floors is 14 cm and live load is 2 kN/m2 according to TS-498. Height
of the stories shows variation only for the first story. Ground story level is 3.5 meters high
and the height of other stories is 3 meters. No basement is considered in the design and the
building is placed directly on fixed supports. Asymmetry and the resulting torsional coupling
are provided by offsetting the mass center of the building from the center of stiffness by 15%
of the plan dimension. Thus, unsymmetrical mass distribution is obtained along the direction
of analysis (Y-axis).
The structure is modeled by using the OpenSees software (2011). Two different models are
prepared for different types of analyses: A linear elastic model is developed for perform-
ing linear elastic response spectrum analysis. A nonlinear structural model is also generated
32
Figure 4.1: Elevation view of the frames in the direction of analysis (all units in meters).
33
in order to perform nonlinear response history analysis, generalized pushover analysis and
modal pushover analysis. In both models, predefined members and material relationships in
the OpenSees platform are utilized. Linear elastic model is composed of “elasticBeamCol-
umn” elements which are defined with geometric properties of the member cross-sections and
shows no material or geometric nonlinearity. In order to account for the cracked section stiff-
ness, gross moments of inertia of the beams are multiplied by 0.5, while that of columns are
multiplied by 0.6.
Due to the complexity and the requirement of high amount of computational effort associ-
ated with the nonlinear analysis, a fully distributed plasticity model is not preferred. Instead,
inelasticity is confined to the member ends. In compliance with this approach, “beamwith-
Hinges” element is utilized in both beams and columns. According to the formulation of this
element, plasticity is “lumped” along a certain hinge length at both ends of the member. Rest
of the element displays linear elastic behavior. These specific hinge lengths are considered as
half of the section depth at all members.
Different plasticity relationships are defined for the aforementioned plasticity zones in beams
and columns. In the case of beams, moment-curvature relationships are defined for each mem-
ber and assigned to the member ends with elasto-plastic hysteresis relations. This approach
is preferred over the more elaborate fiber sections for two major reasons. First, axial loads on
beams show little or no variation during seismic response and remains close to zero. There-
fore a complex section where the moment capacity is updated according to the axial load
variation is unnecessary. Second, when rigid floor diaphragms and fiber sections in beams are
used together, unrealistically high axial loads are observed along the beams. This is a known
issue in OpenSees platform and it is explained by the program authors such that constraining
the neutral axis of a fiber section by a rigid diaphragm results in this type of response.
For each beam cross section, moment-curvature relationships are defined according to the
section detailing. Since demands occurring on each beam along different frames at each story
are different, a single detailing for all beams is not utilized. Instead, beam design is performed
according to the demand on each beam, which resulted in different beam cross-sections for the
structure. This variability in different beam cross-sections is taken into the account while cal-
culating the moment-curvature relationships and hysteresis models are assigned to the beam
ends accordingly.
34
“Hysteretic Material” model is utilized for the hysteresis relationship used in the beams. This
is a bilinear model where moment-curvature relationship is defined. No pinching or deterio-
ration effects are included in the model. For beams, “beamwithHinges” elements are formed
by combining these hysteresis models with the linear elastic properties of the section for the
rest of the member.
Fiber sections are selected in order to introduce plasticity into the columns. The main rea-
son for this choice is that the change in axial load during seismic response is significant in
columns and the bending moment capacity is affected from this change considerably. This
behavior cannot be modeled accurately by defining a moment-curvature relationship obtained
at a single axial load value. Fiber section is a good choice for the described situation since the
response of the section is calculated directly from the material properties and the loads acting
on the member at that instant.
Realistic material models for both reinforcement and concrete are included in the formula-
tion of fiber sections. For reinforcement steel, “Steel01” material model is utilized. It is a
bilinear material model with a very small strain-hardening slope. This behavior is not only
accurate enough, but also relatively simple to represent the reinforcing steel behavior. In the
case of concrete, “Concrete01” model is employed. The model includes a force-deformation
relationship (stress-strain in this case) based on the Modified Kent and Park model (Kent and
Park, 1971) with zero tensile strength. Different material relationships are generated for cover
and core concrete. In the case of core concrete, compressive strength of the confined concrete
(core) is multiplied with Kc which is the amplification factor for confined concrete calculated
according to the aforementioned material model. By selecting such an approach, confinement
effects which occur on the column sections could also be imposed on the analytical model.
Similar to what has been done in the modeling of beams, linear elastic part of the columns are
generated by using the cross-section properties
In the elastic portions of the “beamwithHinges” elements of the nonlinear model, cracked
section stiffness is also considered by following the approach implemented in creating the
linear elastic model. Moment-curvature relationships for the beams and fiber sections for
columns implicitly include this cracking phenomenon and reduction in stiffness; therefore no
extra work is carried out to include these effects in the lumped plasticity regions.
P-Delta effects are also considered in the analytical models. Rather than using linear geomet-
35
ric transformations, P-Delta effects are included in order to obtain a more realistic response.
Rayleigh damping is used in both models where the damping coefficients are obtained from
the 1’st and 3’rd modes. 5% damping is imposed in the solution process for the 1’st and 3’rd
modes.
In order to utilize the described analytical models in various analyses for generating response
results, an automation process is written in MATLAB (R2010b) where response history,
response spectrum, modal pushover and generalized pushover analyses are conducted for the
ground motion set given in Chapter 3. Entire analysis procedures for these methods along with
the pre and post-processing codes are also written in MATLAB. Custom scripts are created
to run OpenSees in conjunction with MATLAB so that the analyses are performed and the
results are produced in a combined procedure.
4.2 Free Vibration Properties
Modal properties of the building are presented in Table 4.1. Among the sixteen torsionally
coupled modes, first nine modes are presented. The first column in Table 4.1 shows the
mode numbers ( X and Y stand for translation dominant mode, and θ stands for the rotation
dominant mode of the Y-θ couple). Corresponding period of each mode is displayed in the
second column. The ordering of modal period indicated that the structure is torsionally stiff.
Effective modal masses (in tons) and the effective modal mass ratios are given in the third and
fourth columns respectively for X and Y directions are presented in the following columns.
4.3 Presentations of the Analysis Results
In the following pages, results compiled from the nonlinear response history analysis (NRHA),
generalized pushover analysis (GPA), modal pushover analysis (MPA) and response history
analysis (RSA) of the eight story structure under the ground motion set consisting of eight
ground motions records described in Chapter 3 are presented. From Figure 4.2 to Figure 4.43,
maximum interstory drift ratios, mean beam plastic rotations, mean column chord rotations,
end moments of beams in the 2’nd and the 6’th stories and story shears of all Y-direction
frames (introduced as FE, FI, SI, SE in Figure 2.2) are plotted. For the first two ground
motions, also presented are the beam plastic rotations of the 2’nd and the 6’th story beams
36
Table 4.1: Free vibration properties of the eight story structure
ModePeriod
(seconds)
EffectiveModal Mass
inY-direction
(M∗n,y) (tons)
EffectiveModal Mass
Ratio inY-direction
EffectiveModal Mass
inX-direction(M∗n,x) (tons)
EffectiveModal Mass
Ratio inX-direction
1X 1.69 0 0 1564 0.8211Y 1.62 1325.28 0.695 0 01θ 1.03 258.81 0.136 0 02Y 0.53 157.47 0.083 0 02X 0.51 0 0 187.91 0.0992θ 0.33 31.21 0.016 0 03Y 0.29 56.69 0.030 0 03X 0.28 0 0 74.22 0.0393θ 0.19 29.58 0.016 0 0
located on these frames.
In general, it is seen that results produced by GPA are satisfactory compared to the NRHA
results which are regarded as the benchmark values. Under moderate to low demand ground
motions, performance of each analysis method is close to each other as can be seen in GM7
and GM8 plots. However, differences increase as seismic demand becomes higher. Despite
increasing differences, GPA mostly yields results sufficiently close to the benchmark values.
GPA is successful in the estimation of higher mode effects especially under GM2, GM3 and
GM4 which excite the second mode. On the other hand, under GM1, GM5 and GM6, some
overshooting in the FE frame results are observed. This is due to the strong demands that
these ground motions impose on the structure which in turn creates high inelastic demands
and increased amounts of nonlinearity in dynamic response. The resulting nonlinearity that
changes the deformation pattern of the structure in dynamic response is difficult to capture
with the static analysis procedures.
For each ground motion, GPA is very successful in terms of estimating the internal forces as
can be seen from the beam end moments and story shears presented in the figures. Since no
modal combination rule is employed in gathering the total response, internal forces can be
estimated almost exactly contrary to the other static analysis methods.
37
0
1
2
3
4
5
6
7
8
0 0,005 0,01 0,015 0,02 0,025 0,03
Story
#
Frame FE, GM1
0
1
2
3
4
5
6
7
8
0 0,005 0,01 0,015 0,02 0,025 0,03
Frame FI, GM1
0
1
2
3
4
5
6
7
8
0 0,005 0,01 0,015 0,02 0,025 0,03
Story
#
Interstory Drift Ratio
Frame SI, GM1
0
1
2
3
4
5
6
7
8
0 0,005 0,01 0,015 0,02 0,025 0,03
Interstory Drift Ratio
Frame SE, GM1
NRHA GPA MPA RSA
Figure 4.2: Interstory drift ratios of four frames under GM1
0
1
2
3
4
5
6
7
8
0 0,005 0,01 0,015 0,02 0,025
Story
#
Frame FE, GM1
0
1
2
3
4
5
6
7
8
0 0,005 0,01 0,015 0,02 0,025
Frame FI, GM1
0
1
2
3
4
5
6
7
8
0 0,005 0,01 0,015 0,02 0,025
Story
#
Plastic Rotation (rad)
Frame SI, GM1
0
1
2
3
4
5
6
7
8
0 0,005 0,01 0,015 0,02 0,025
Plastic Rotation (rad)
Frame SE, GM1
NRHA GPA MPA
Figure 4.3: Mean beam plastic rotations of four frames under GM1
38
Second Story Beam Plastic Rotations
Sixth Story Beam Plastic Rotations
0
0,005
0,01
0,015
0,02
0,025
Pla
stic
Ro
tati
on
(ra
d)
Frame FE, GM1
0
0,005
0,01
0,015
0,02
0,025Frame FI, GM1
0
0,005
0,01
0,015
0,02
0,025
Pla
stic
Ro
tati
on
(ra
d)
Frame SI, GM1
0
0,005
0,01
0,015
0,02
0,025Frame SE, GM1
0
0,001
0,002
0,003
0,004
0,005
0,006
0,007
Pla
stic
Ro
tati
on
(ra
d)
Frame FE, GM1
0
0,001
0,002
0,003
0,004
0,005
0,006
0,007
Frame FI, GM1
0
0,001
0,002
0,003
0,004
0,005
0,006
0,007
Pla
stic
Ro
tati
on
(ra
d)
Frame SI, GM1
0
0,001
0,002
0,003
0,004
0,005
0,006
0,007Frame SE, GM1
NRHA GPA MPA
Exterior Beam Exterior Beam Interior
Beam Exterior Beam Exterior Beam Interior
Beam
Exterior Beam Exterior Beam
Interior
Beam Exterior Beam Exterior Beam
Interior
Beam
Figure 4.4: 2’nd and 6’th story beam plastic rotations of four frames under GM1
39
I Ends! J Ends!
0
1
2
3
4
5
6
7
8
0 0,005 0,01 0,015
Story
!#
Frame!FE,!GM1
0
1
2
3
4
5
6
7
8
0 0,0025 0,005 0,0075 0,01
Frame!FE,!GM1
0
1
2
3
4
5
6
7
8
0 0,005 0,01 0,015
Story
!#
Frame!FI,!GM1
0
1
2
3
4
5
6
7
8
0 0,0025 0,005 0,0075 0,01
Frame!FI,!GM1
0
1
2
3
4
5
6
7
8
0 0,005 0,01 0,015
Story
!#
Frame!SI,!GM1
0
1
2
3
4
5
6
7
8
0 0,0025 0,005 0,0075 0,01
Frame!SI,!GM1
0
1
2
3
4
5
6
7
8
0 0,005 0,01 0,015
Story
!#
Chord!Rotation!(rad)
Frame!SE,!GM1
0
1
2
3
4
5
6
7
8
0 0,0025 0,005 0,0075 0,01
Chord!Rotation!(rad)
Frame!SE,!GM1
NRHA GPA MPA RSA
Figure 4.5: Mean column chord rotations of four frames under GM1
40
Second Story Beam End Moments
Sixth Story Beam End Moments
0
100
200
300
400
Moment (kNm)
Frame FE, GM1
0
100
200
300
400
Frame FI, GM1
0
100
200
300
400
Moment (kNm)
Frame SI, GM1
0
100
200
300
400
Frame SE, GM1
0
100
200
300
400
Moment (kNm)
Frame FE, GM1
0
100
200
300
400
Frame FI, GM1
0
100
200
300
400
Moment (kNm)
Frame SI, GM1
0
100
200
300
400
Frame SE, GM1
NRHA GPA MPA
Exterior Beam Exterior Beam
Interior
Beam Exterior Beam Exterior Beam
Interior
Beam
Exterior Beam Exterior Beam
Interior
Beam Exterior Beam Exterior Beam
Interior
Beam
Figure 4.6: 2’nd and 6’th story beam end moments of four frames under GM1
41
0
1
2
3
4
5
6
7
8
0 150 300 450 600 750 900
Story
#
Frame FE, GM1
0
1
2
3
4
5
6
7
8
0 150 300 450 600 750 900
Frame FI, GM1
0
1
2
3
4
5
6
7
8
0 150 300 450 600 750 900
Story
#
Shear Force (kN)
Frame SI, GM1
0
1
2
3
4
5
6
7
8
0 150 300 450 600 750 900
Shear Force (kN)
Frame SE, GM1
NRHA GPA MPA
Figure 4.7: Story shear forces of four frames under GM1
42
0
1
2
3
4
5
6
7
8
0 0,005 0,01 0,015 0,02
Story
#
Frame FE, GM2
0
1
2
3
4
5
6
7
8
0 0,005 0,01 0,015 0,02
Frame FI, GM2
0
1
2
3
4
5
6
7
8
0 0,005 0,01 0,015 0,02
Story
#
Interstory Drift Ratio
Frame SI, GM2
0
1
2
3
4
5
6
7
8
0 0,005 0,01 0,015 0,02
Interstory Drift Ratio
Frame SE, GM2
NRHA GPA MPA RSA
Figure 4.8: Interstory drift ratios of four frames under GM2
0
1
2
3
4
5
6
7
8
0 0,005 0,01 0,015
Story
#
Frame FE, GM2
0
1
2
3
4
5
6
7
8
0 0,005 0,01 0,015
Frame FI, GM2
0
1
2
3
4
5
6
7
8
0 0,005 0,01 0,015
Story
#
Plastic Rotation (rad)
Frame SI, GM2
0
1
2
3
4
5
6
7
8
0 0,005 0,01 0,015
Plastic Rotation (rad)
Frame SE, GM2
NRHA GPA MPA
Figure 4.9: Mean beam plastic rotations of four frames under GM2
43
Second Story Beam Plastic Rotations
Sixth Story Beam Plastic Rotations
0
0,002
0,004
0,006
0,008
0,01
0,012
Pla
stic
Ro
tati
on
(ra
d)
Frame FE, GM2
0
0,002
0,004
0,006
0,008
0,01
0,012Frame FI, GM2
0
0,002
0,004
0,006
0,008
0,01
0,012
Pla
stic
Ro
tati
on
(ra
d)
Frame SI, GM2
0
0,002
0,004
0,006
0,008
0,01
0,012Frame SE, GM2
0
0,002
0,004
0,006
0,008
0,01
0,012
Pla
stic
Ro
tati
on
(ra
d)
Frame FE, GM2
0
0,002
0,004
0,006
0,008
0,01
0,012Frame FI, GM2
0
0,002
0,004
0,006
0,008
0,01
0,012
Pla
stic
Ro
tati
on
(ra
d)
Frame SI, GM2
0
0,002
0,004
0,006
0,008
0,01
0,012Frame SE, GM2
NRHA GPA MPA
Exterior Beam Exterior Beam Interior
Beam Exterior Beam Exterior Beam Interior
Beam
Exterior Beam Exterior Beam
Interior
Beam Exterior Beam Exterior Beam
Interior
Beam
Figure 4.10: 2’nd and 6’th story beam plastic rotations of four frames under GM2
44
I Ends! J Ends!
0
1
2
3
4
5
6
7
8
0 0,0025 0,005 0,0075 0,01 0,0125
Story
!#
Frame!FE,!GM2
0
1
2
3
4
5
6
7
8
0 0,0025 0,005 0,0075
Frame!FE,!GM2
0
1
2
3
4
5
6
7
8
0 0,0025 0,005 0,0075 0,01 0,0125
Story
!#
Frame!FI,!GM2
0
1
2
3
4
5
6
7
8
0 0,0025 0,005 0,0075
Frame!FI,!GM2
0
1
2
3
4
5
6
7
8
0 0,0025 0,005 0,0075 0,01 0,0125
Story
!#
Frame!SI,!GM2
0
1
2
3
4
5
6
7
8
0 0,0025 0,005 0,0075
Frame!SI,!GM2
0
1
2
3
4
5
6
7
8
0 0,0025 0,005 0,0075 0,01 0,0125
Story
!#
Chord!Rotation!(rad)
Frame!SE,!GM2
0
1
2
3
4
5
6
7
8
0 0,0025 0,005 0,0075
Chord!Rotation!(rad)
Frame!SE,!GM2
NRHA GPA MPA RSA
Figure 4.11: Mean column chord rotations of four frames under GM2
45
Second Story Beam End Moments
Sixth Story Beam End Moments
0
100
200
300
400
Moment (kNm)
Frame FE, GM2
0
100
200
300
400
Frame FI, GM2
0
100
200
300
400
Moment (kNm)
Frame SI, GM2
0
100
200
300
400
Frame SE, GM2
0
100
200
300
400
Moment (kNm)
Frame FE, GM2
0
100
200
300
400
Frame FI, GM2
0
100
200
300
400
Moment (kNm)
Frame SI, GM2
0
100
200
300
400
Frame SE, GM2
NRHA GPA MPA
Exterior Beam Exterior Beam
Interior
Beam Exterior Beam Exterior Beam
Interior
Beam
Exterior Beam Exterior Beam
Interior
Beam Exterior Beam Exterior Beam
Interior
Beam
Figure 4.12: 2’nd and 6’th story beam end moments of four frames under GM2
46
0
1
2
3
4
5
6
7
8
0 200 400 600 800 1000
Story
#
Frame FE, GM2
0
1
2
3
4
5
6
7
8
0 200 400 600 800 1000
Frame FI, GM2
0
1
2
3
4
5
6
7
8
0 200 400 600 800 1000
Story
#
Shear Force (kN)
Frame SI, GM2
0
1
2
3
4
5
6
7
8
0 200 400 600 800 1000
Shear Force (kN)
Frame SE, GM2
NRHA GPA MPA
Figure 4.13: Story shear forces of four frames under GM2
47
0
1
2
3
4
5
6
7
8
0 0,005 0,01 0,015 0,02 0,025 0,03
Story
#
Frame FE, GM3
0
1
2
3
4
5
6
7
8
0 0,005 0,01 0,015 0,02 0,025 0,03
Frame FI, GM3
0
1
2
3
4
5
6
7
8
0 0,005 0,01 0,015 0,02 0,025 0,03
Story
#
Interstory Drift Ratio
Frame SI, GM3
0
1
2
3
4
5
6
7
8
0 0,005 0,01 0,015 0,02 0,025 0,03
Interstory Drift Ratio
Frame SE, GM3
NRHA GPA MPA RSA
Figure 4.14: Interstory drift ratios of four frames under GM3
0
1
2
3
4
5
6
7
8
0 0,005 0,01 0,015 0,02 0,025
Story
#
Frame FE, GM3
0
1
2
3
4
5
6
7
8
0 0,005 0,01 0,015 0,02 0,025
Frame FI, GM3
0
1
2
3
4
5
6
7
8
0 0,005 0,01 0,015 0,02 0,025
Story
#
Plastic Rotation (rad)
Frame SI, GM3
0
1
2
3
4
5
6
7
8
0 0,005 0,01 0,015 0,02 0,025
Plastic Rotation (rad)
Frame SE, GM3
NRHA GPA MPA
Figure 4.15: Mean beam plastic rotations of four frames under GM3
48
I Ends! J Ends!
0
1
2
3
4
5
6
7
8
0 0,005 0,01 0,015 0,02 0,025
Story
!#
Frame!FE,!GM3
0
1
2
3
4
5
6
7
8
0 0,0025 0,005 0,0075 0,01 0,0125 0,015
Frame!FE,!GM3
0
1
2
3
4
5
6
7
8
0 0,005 0,01 0,015 0,02 0,025
Story
!#
Frame!FI,!GM3
0
1
2
3
4
5
6
7
8
0 0,0025 0,005 0,0075 0,01 0,0125 0,015
Frame!FI,!GM3
0
1
2
3
4
5
6
7
8
0 0,005 0,01 0,015 0,02 0,025
Story
!#
Frame!SI,!GM3
0
1
2
3
4
5
6
7
8
0 0,0025 0,005 0,0075 0,01 0,0125 0,015
Frame!SI,!GM3
0
1
2
3
4
5
6
7
8
0 0,005 0,01 0,015 0,02 0,025
Story
!#
Chord!Rotation!(rad)
Frame!SE,!GM3
0
1
2
3
4
5
6
7
8
0 0,0025 0,005 0,0075 0,01 0,0125 0,015
Chord!Rotation!(rad)
Frame!SE,!GM3
NRHA GPA MPA RSA
Figure 4.16: Mean column chord rotations of four frames under GM3
49
Second Story Beam End Moments
Sixth Story Beam End Moments
0
100
200
300
400
Moment (kNm)
Frame FE, GM3
0
100
200
300
400
Frame FI, GM3
0
100
200
300
400
Moment (kNm)
Frame SI, GM3
0
100
200
300
400
Frame SE, GM3
0
100
200
300
400
Moment (kNm)
Frame FE, GM3
0
100
200
300
400
Frame FI, GM3
0
100
200
300
400
Moment (kNm)
Frame SI, GM3
0
100
200
300
400
Frame SE, GM3
NRHA GPA MPA
Exterior Beam Exterior Beam
Interior
Beam Exterior Beam Exterior Beam
Interior
Beam
Exterior Beam Exterior Beam
Interior
Beam Exterior Beam Exterior Beam
Interior
Beam
Figure 4.17: 2’nd and 6’th story beam end moments of four frames under GM3
50
0
1
2
3
4
5
6
7
8
0 200 400 600 800 1000
Story
#
Frame FE, GM3
0
1
2
3
4
5
6
7
8
0 200 400 600 800 1000
Frame FI, GM3
0
1
2
3
4
5
6
7
8
0 200 400 600 800 1000
Story
#
Shear Force (kN)
Frame SI, GM3
0
1
2
3
4
5
6
7
8
0 200 400 600 800 1000
Shear Force (kN)
Frame SE, GM3
NRHA GPA MPA
Figure 4.18: Story shear forces of four frames under GM3
51
0
1
2
3
4
5
6
7
8
0 0,005 0,01 0,015 0,02 0,025 0,03
Story
#
Frame FE, GM4
0
1
2
3
4
5
6
7
8
0 0,005 0,01 0,015 0,02 0,025 0,03
Frame FI, GM4
0
1
2
3
4
5
6
7
8
0 0,005 0,01 0,015 0,02 0,025 0,03
Story
#
Interstory Drift Ratio
Frame SI, GM4
0
1
2
3
4
5
6
7
8
0 0,005 0,01 0,015 0,02 0,025 0,03
Interstory Drift Ratio
Frame SE, GM4
NRHA GPA MPA RSA
Figure 4.19: Interstory drift ratios of four frames under GM4
0
1
2
3
4
5
6
7
8
0 0,005 0,01 0,015 0,02 0,025
Story
#
Frame FE, GM4
0
1
2
3
4
5
6
7
8
0 0,005 0,01 0,015 0,02 0,025
Frame FI, GM4
0
1
2
3
4
5
6
7
8
0 0,005 0,01 0,015 0,02 0,025
Story
#
Plastic Rotation (rad)
Frame SI, GM4
0
1
2
3
4
5
6
7
8
0 0,005 0,01 0,015 0,02 0,025
Plastic Rotation (rad)
Frame SE, GM4
NRHA GPA MPA
Figure 4.20: Mean beam plastic rotations of four frames under GM4
52
I Ends! J Ends!
0
1
2
3
4
5
6
7
8
0 0,005 0,01 0,015 0,02 0,025
Story
!#
Frame!FE,!GM4
0
1
2
3
4
5
6
7
8
0 0,0025 0,005 0,0075 0,01 0,0125 0,015
Frame!FE,!GM4
0
1
2
3
4
5
6
7
8
0 0,005 0,01 0,015 0,02 0,025
Story
!#
Frame!FI,!GM4
0
1
2
3
4
5
6
7
8
0 0,0025 0,005 0,0075 0,01 0,0125 0,015
Frame!FI,!GM4
0
1
2
3
4
5
6
7
8
0 0,005 0,01 0,015 0,02 0,025
Story
!#
Frame!SI,!GM4
0
1
2
3
4
5
6
7
8
0 0,0025 0,005 0,0075 0,01 0,0125 0,015
Frame!SI,!GM4
0
1
2
3
4
5
6
7
8
0 0,005 0,01 0,015 0,02 0,025
Story
!#
Chord!Rotation!(rad)
Frame!SE,!GM4
0
1
2
3
4
5
6
7
8
0 0,0025 0,005 0,0075 0,01 0,0125 0,015
Chord!Rotation!(rad)
Frame!SE,!GM4
NRHA GPA MPA RSA
Figure 4.21: Mean column chord rotations of four frames under GM4
53
Second Story Beam End Moments
Sixth Story Beam End Moments
0
100
200
300
400
Moment (kNm)
Frame FE, GM4
0
100
200
300
400
Frame FI, GM4
0
100
200
300
400
Moment (kNm)
Frame SI, GM4
0
100
200
300
400
Frame SE, GM4
0
100
200
300
400
Moment (kNm)
Frame FE, GM4
0
100
200
300
400
Frame FI, GM4
0
100
200
300
400
Moment (kNm)
Frame SI, GM4
0
100
200
300
400
Frame SE, GM4
NRHA GPA MPA
Exterior Beam Exterior Beam
Interior
Beam Exterior Beam Exterior Beam
Interior
Beam
Exterior Beam Exterior Beam
Interior
Beam Exterior Beam Exterior Beam
Interior
Beam
Figure 4.22: 2’nd and 6’th story beam end moments of four frames under GM4
54
0
1
2
3
4
5
6
7
8
0 200 400 600 800 1000
Story
#
Frame FE, GM4
0
1
2
3
4
5
6
7
8
0 200 400 600 800 1000
Frame FI, GM4
0
1
2
3
4
5
6
7
8
0 200 400 600 800 1000
Story
#
Shear Force (kN)
Frame SI, GM4
0
1
2
3
4
5
6
7
8
0 200 400 600 800 1000
Shear Force (kN)
Frame SE, GM4
NRHA GPA MPA
Figure 4.23: Story shear forces of four frames under GM4
55
0
1
2
3
4
5
6
7
8
0 0,005 0,01 0,015 0,02 0,025 0,03
Story
#
Frame FE, GM5
0
1
2
3
4
5
6
7
8
0 0,005 0,01 0,015 0,02 0,025 0,03
Frame FI, GM5
0
1
2
3
4
5
6
7
8
0 0,005 0,01 0,015 0,02 0,025 0,03
Story
#
Interstory Drift Ratio
Frame SI, GM5
0
1
2
3
4
5
6
7
8
0 0,005 0,01 0,015 0,02 0,025 0,03
Interstory Drift Ratio
Frame SE, GM5
NRHA GPA MPA RSA
Figure 4.24: Interstory drift ratios of four frames under GM5
0
1
2
3
4
5
6
7
8
0 0,005 0,01 0,015 0,02 0,025
Story
#
Frame FE, GM5
0
1
2
3
4
5
6
7
8
0 0,005 0,01 0,015 0,02 0,025
Frame FI, GM5
0
1
2
3
4
5
6
7
8
0 0,005 0,01 0,015 0,02 0,025
Story
#
Plastic Rotation (rad)
Frame SI, GM5
0
1
2
3
4
5
6
7
8
0 0,005 0,01 0,015 0,02 0,025
Plastic Rotation (rad)
Frame SE, GM5
NRHA GPA MPA
Figure 4.25: Mean beam plastic rotations of four frames under GM5
56
I Ends! J Ends!
0
1
2
3
4
5
6
7
8
0 0,005 0,01 0,015 0,02
Story
!#
Frame!FE,!GM5
0
1
2
3
4
5
6
7
8
0 0,0025 0,005 0,0075 0,01
Frame!FE,!GM5
0
1
2
3
4
5
6
7
8
0 0,005 0,01 0,015 0,02
Story
!#
Frame!FI,!GM5
0
1
2
3
4
5
6
7
8
0 0,0025 0,005 0,0075 0,01
Frame!FI,!GM5
0
1
2
3
4
5
6
7
8
0 0,005 0,01 0,015 0,02
Story
!#
Frame!SI,!GM5
0
1
2
3
4
5
6
7
8
0 0,0025 0,005 0,0075 0,01
Frame!SI,!GM5
0
1
2
3
4
5
6
7
8
0 0,005 0,01 0,015 0,02
Story
!#
Chord!Rotation!(rad)
Frame!SE,!GM5
0
1
2
3
4
5
6
7
8
0 0,0025 0,005 0,0075 0,01
Chord!Rotation!(rad)
Frame!SE,!GM5
NRHA GPA MPA RSA
Figure 4.26: Mean column chord rotations of four frames under GM5
57
Second Story Beam End Moments
Sixth Story Beam End Moments
0
100
200
300
400
Moment (kNm)
Frame FE, GM5
0
100
200
300
400
Frame FI, GM5
0
100
200
300
400
Moment (kNm)
Frame SI, GM5
0
100
200
300
400
Frame SE, GM5
0
100
200
300
400
Moment (kNm)
Frame FE, GM5
0
100
200
300
400
Frame FI, GM5
0
100
200
300
400
Moment (kNm)
Frame SI, GM5
0
100
200
300
400
Frame SE, GM5
NRHA GPA MPA
Exterior Beam Exterior Beam
Interior
Beam Exterior Beam Exterior Beam
Interior
Beam
Exterior Beam Exterior Beam
Interior
Beam Exterior Beam Exterior Beam
Interior
Beam
Figure 4.27: 2’nd and 6’th story beam end moments of four frames under GM5
58
0
1
2
3
4
5
6
7
8
0 200 400 600 800 1000
Story
#
Frame FE, GM5
0
1
2
3
4
5
6
7
8
0 200 400 600 800 1000
Frame FI, GM5
0
1
2
3
4
5
6
7
8
0 200 400 600 800 1000
Story
#
Shear Force (kN)
Frame SI, GM5
0
1
2
3
4
5
6
7
8
0 200 400 600 800 1000
Shear Force (kN)
Frame SE, GM5
NRHA GPA MPA
Figure 4.28: Story shear forces of four frames under GM5
59
0
1
2
3
4
5
6
7
8
0 0,005 0,01 0,015 0,02 0,025 0,03
Story
#
Frame FE, GM6
0
1
2
3
4
5
6
7
8
0 0,005 0,01 0,015 0,02 0,025 0,03
Frame FI, GM6
0
1
2
3
4
5
6
7
8
0 0,005 0,01 0,015 0,02 0,025 0,03
Story
#
Interstory Drift Ratio
Frame SI, GM6
0
1
2
3
4
5
6
7
8
0 0,005 0,01 0,015 0,02 0,025 0,03
Interstory Drift Ratio
Frame SE, GM6
NRHA GPA MPA RSA
Figure 4.29: Interstory drift ratios of four frames under GM6
0
1
2
3
4
5
6
7
8
0 0,005 0,01 0,015 0,02 0,025
Story
#
Frame FE, GM6
0
1
2
3
4
5
6
7
8
0 0,005 0,01 0,015 0,02 0,025
Frame FI, GM6
0
1
2
3
4
5
6
7
8
0 0,005 0,01 0,015 0,02 0,025
Story
#
Plastic Rotation (rad)
Frame SI, GM6
0
1
2
3
4
5
6
7
8
0 0,005 0,01 0,015 0,02 0,025
Plastic Rotation (rad)
Frame SE, GM6
NRHA GPA MPA
Figure 4.30: Mean beam plastic rotations of four frames under GM6
60
I Ends! J Ends!
0
1
2
3
4
5
6
7
8
0 0,005 0,01 0,015
Story
!#
Frame!FE,!GM6
0
1
2
3
4
5
6
7
8
0 0,0015 0,003 0,0045 0,006 0,0075
Frame!FE,!GM6
0
1
2
3
4
5
6
7
8
0 0,005 0,01 0,015
Story
!#
Frame!FI,!GM6
0
1
2
3
4
5
6
7
8
0 0,0015 0,003 0,0045 0,006 0,0075
Frame!FI,!GM6
0
1
2
3
4
5
6
7
8
0 0,005 0,01 0,015
Story
!#
Frame!SI,!GM6
0
1
2
3
4
5
6
7
8
0 0,0015 0,003 0,0045 0,006 0,0075
Frame!SI,!GM6
0
1
2
3
4
5
6
7
8
0 0,005 0,01 0,015
Story
!#
Chord!Rotation!(rad)
Frame!SE,!GM6
0
1
2
3
4
5
6
7
8
0 0,0015 0,003 0,0045 0,006 0,0075
Chord!Rotation!(rad)
Frame!SE,!GM6
NRHA GPA MPA RSA
Figure 4.31: Mean column chord rotations of four frames under GM6
61
Second Story Beam End Moments
Sixth Story Beam End Moments
0
100
200
300
400
Moment (kNm)
Frame FE, GM6
0
100
200
300
400
Frame FI, GM6
0
100
200
300
400
Moment (kNm)
Frame SI, GM6
0
100
200
300
400
Frame SE, GM6
0
100
200
300
400
Moment (kNm)
Frame FE, GM6
0
100
200
300
400
Frame FI, GM6
0
100
200
300
400
Moment (kNm)
Frame SI, GM6
0
100
200
300
400
Frame SE, GM6
NRHA GPA MPA
Exterior Beam Exterior Beam
Interior
Beam Exterior Beam Exterior Beam
Interior
Beam
Exterior Beam Exterior Beam
Interior
Beam Exterior Beam Exterior Beam
Interior
Beam
Figure 4.32: 2’nd and 6’th story beam end moments of four frames under GM6
62
0
1
2
3
4
5
6
7
8
0 200 400 600 800
Story
#
Frame FE, GM6
0
1
2
3
4
5
6
7
8
0 200 400 600 800
Frame FI, GM6
0
1
2
3
4
5
6
7
8
0 200 400 600 800
Story
#
Shear Force (kN)
Frame SI, GM6
0
1
2
3
4
5
6
7
8
0 200 400 600 800
Shear Force (kN)
Frame SE, GM6
NRHA GPA MPA
Figure 4.33: Story shear forces of four frames under GM6
63
0
1
2
3
4
5
6
7
8
0 0,005 0,01 0,015 0,02 0,025
Story
#
Frame FE, GM7
0
1
2
3
4
5
6
7
8
0 0,005 0,01 0,015 0,02 0,025
Frame FI, GM7
0
1
2
3
4
5
6
7
8
0 0,005 0,01 0,015 0,02 0,025
Story
#
Interstory Drift Ratio
Frame SI, GM7
0
1
2
3
4
5
6
7
8
0 0,005 0,01 0,015 0,02 0,025
Interstory Drift Ratio
Frame SE, GM7
NRHA GPA MPA RSA
Figure 4.34: Interstory drift ratios of four frames under GM7
0
1
2
3
4
5
6
7
8
0 0,005 0,01 0,015 0,02
Story
#
Frame FE, GM7
0
1
2
3
4
5
6
7
8
0 0,005 0,01 0,015 0,02
Frame FI, GM7
0
1
2
3
4
5
6
7
8
0 0,005 0,01 0,015 0,02
Story
#
Plastic Rotation (rad)
Frame SI, GM7
0
1
2
3
4
5
6
7
8
0 0,005 0,01 0,015 0,02
Plastic Rotation (rad)
Frame SE, GM7
NRHA GPA MPA
Figure 4.35: Mean beam plastic rotations of four frames under GM7
64
I Ends! J Ends!
0
1
2
3
4
5
6
7
8
0 0,005 0,01 0,015
Story
!#
Frame!FE,!GM7
0
1
2
3
4
5
6
7
8
0 0,0015 0,003 0,0045 0,006 0,0075
Frame!FE,!GM7
0
1
2
3
4
5
6
7
8
0 0,005 0,01 0,015
Story
!#
Frame!FI,!GM7
0
1
2
3
4
5
6
7
8
0 0,0015 0,003 0,0045 0,006 0,0075
Frame!FI,!GM7
0
1
2
3
4
5
6
7
8
0 0,005 0,01 0,015
Story
!#
Frame!SI,!GM7
0
1
2
3
4
5
6
7
8
0 0,0015 0,003 0,0045 0,006 0,0075
Frame!SI,!GM7
0
1
2
3
4
5
6
7
8
0 0,005 0,01 0,015
Story
!#
Chord!Rotation!(rad)
Frame!SE,!GM7
0
1
2
3
4
5
6
7
8
0 0,0015 0,003 0,0045 0,006 0,0075
Chord!Rotation!(rad)
Frame!SE,!GM7
NRHA GPA MPA RSA
Figure 4.36: Mean column chord rotations of four frames under GM7
65
Second Story Beam End Moments
Sixth Story Beam End Moments
0
100
200
300
400
Moment (kNm)
Frame FE, GM7
0
100
200
300
400
Frame FI, GM7
0
100
200
300
400
Moment (kNm)
Frame SI, GM7
0
100
200
300
400
Frame SE, GM7
0
100
200
300
400
Moment (kNm)
Frame FE, GM7
0
100
200
300
400
Frame FI, GM7
0
100
200
300
400
Moment (kNm)
Frame SI, GM7
0
100
200
300
400
Frame SE, GM7
NRHA GPA MPA
Exterior Beam Exterior Beam
Interior
Beam Exterior Beam Exterior Beam
Interior
Beam
Exterior Beam Exterior Beam
Interior
Beam Exterior Beam Exterior Beam
Interior
Beam
Figure 4.37: 2’nd and 6’th story beam end moments of four frames under GM7
66
0
1
2
3
4
5
6
7
8
0 200 400 600 800
Story
#
Frame FE, GM7
0
1
2
3
4
5
6
7
8
0 200 400 600 800
Frame FI, GM7
0
1
2
3
4
5
6
7
8
0 200 400 600 800
Story
#
Shear Force (kN)
Frame SI, GM7
0
1
2
3
4
5
6
7
8
0 200 400 600 800
Shear Force (kN)
Frame SE, GM7
NRHA GPA MPA
Figure 4.38: Story shear forces of four frames under GM7
67
0
1
2
3
4
5
6
7
8
0 0,005 0,01 0,015 0,02 0,025
Story
#
Frame FE, GM8
0
1
2
3
4
5
6
7
8
0 0,005 0,01 0,015 0,02 0,025
Frame FI, GM8
0
1
2
3
4
5
6
7
8
0 0,005 0,01 0,015 0,02 0,025
Story
#
Interstory Drift Ratio
Frame SI, GM8
0
1
2
3
4
5
6
7
8
0 0,005 0,01 0,015 0,02 0,025
Interstory Drift Ratio
Frame SE, GM8
NRHA GPA MPA RSA
Figure 4.39: Interstory drift ratios of four frames under GM8
0
1
2
3
4
5
6
7
8
0 0,005 0,01 0,015 0,02
Story
#
Frame FE, GM8
0
1
2
3
4
5
6
7
8
0 0,005 0,01 0,015 0,02
Frame FI, GM8
0
1
2
3
4
5
6
7
8
0 0,005 0,01 0,015 0,02
Story
#
Plastic Rotation (rad)
Frame SI, GM8
0
1
2
3
4
5
6
7
8
0 0,005 0,01 0,015 0,02
Plastic Rotation (rad)
Frame SE, GM8
NRHA GPA MPA
Figure 4.40: Mean beam plastic rotations of four frames under GM8
68
I Ends! J Ends!
0
1
2
3
4
5
6
7
8
0 0,005 0,01 0,015 0,02
Story
!#
Frame!FE,!GM8
0
1
2
3
4
5
6
7
8
0 0,0015 0,003 0,0045 0,006 0,0075
Frame!FE,!GM8
0
1
2
3
4
5
6
7
8
0 0,005 0,01 0,015 0,02
Story
!#
Frame!FI,!GM8
0
1
2
3
4
5
6
7
8
0 0,0015 0,003 0,0045 0,006 0,0075
Frame!FI,!GM8
0
1
2
3
4
5
6
7
8
0 0,005 0,01 0,015 0,02
Story
!#
Frame!SI,!GM8
0
1
2
3
4
5
6
7
8
0 0,0015 0,003 0,0045 0,006 0,0075
Frame!SI,!GM8
0
1
2
3
4
5
6
7
8
0 0,005 0,01 0,015 0,02
Story
!#
Chord!Rotation!(rad)
Frame!SE,!GM8
0
1
2
3
4
5
6
7
8
0 0,0015 0,003 0,0045 0,006 0,0075
Chord!Rotation!(rad)
Frame!SE,!GM8
NRHA GPA MPA RSA
Figure 4.41: Mean column chord rotations of four frames under GM8
69
Second Story Beam End Moments
Sixth Story Beam End Moments
0
100
200
300
400
Moment (kNm)
Frame FE, GM8
0
100
200
300
400
Frame FI, GM8
0
100
200
300
400
Moment (kNm)
Frame SI, GM8
0
100
200
300
400
Frame SE, GM8
0
100
200
300
400
Moment (kNm)
Frame FE, GM8
0
100
200
300
400
Frame FI, GM8
0
100
200
300
400
Moment (kNm)
Frame SI, GM8
0
100
200
300
400
Frame SE, GM8
NRHA GPA MPA
Exterior Beam Exterior Beam
Interior
Beam Exterior Beam Exterior Beam
Interior
Beam
Exterior Beam Exterior Beam
Interior
Beam Exterior Beam Exterior Beam
Interior
Beam
Figure 4.42: 2’nd and 6’th story beam end moments of four frames under GM8
70
0
1
2
3
4
5
6
7
8
0 200 400 600 800
Story
#
Frame FE, GM8
0
1
2
3
4
5
6
7
8
0 200 400 600 800
Frame FI, GM8
0
1
2
3
4
5
6
7
8
0 200 400 600 800
Story
#
Shear Force (kN)
Frame SI, GM8
0
1
2
3
4
5
6
7
8
0 200 400 600 800
Shear Force (kN)
Frame SE, GM8
NRHA GPA MPA
Figure 4.43: Story shear forces of four frames under GM8
71
CHAPTER 5
CASE STUDY II: TWENTY STORY
UNSYMMETRICAL-PLAN BUILDING
5.1 General Information and Modeling of the Building
The second case study is a twenty story unsymmetrical-plan reinforced concrete building. It
is a tall structure where a “C” shaped core is designed to resist most of the lateral forces.
Different from the previous case study where a shift in the mass center has been introduced,
asymmetry in the direction of analysis (y-axis) is imposed by placing the C-core in an offset
position from the mass center. Due to asymmetric nature of the shape of the core and its
position in the plan, stiffness center shifts from the mass center of the building and torsional
coupling develops under horizontal excitation in the y direction.
Plan of the building and the elevation views of the frames in the direction of analysis are
given in Figure 5.1 and Figure 5.2, respectively. TS-500 and TEC are the main guidelines in
the design where capacity design principles are attained similar to the design of the previous
case study. Seismicity and soil conditions assumed in design are seismic zone 1 with local
site class Z3 in TEC. The enhanced ductility level for frame-wall systems (where moment
frames and shear walls are used together to resist the lateral loads) which corresponds to R=7
is intended in the preliminary design. Load reduction factor R has later been revised to 6.87
in the direction of analysis according to TEC, Section 2.5.2.
As defined by the structural layout, all beams have 8 meter spans. These large spans resulted
in a challenging design process where high strength material properties and large section
dimensions has been selected. C50 grade concrete and S420 steel is employed in reinforced
72
Figu
re5.
1:Pl
anvi
ewof
the
twen
ty-s
tory
unsy
mm
etri
cal-
plan
stru
ctur
e(a
llun
itsin
met
ers)
.
73
Figure 5.2: Elevation view of the frames in the direction of analysis (all units in meters).
74
concrete design. Columns are grouped into two as interior and exterior columns, and two
column sections have been designed accordingly. Interior columns are 70x70 cm2 and exterior
columns are 55x55 cm2 throughout the building. Beams are selected as 30x55 cm2 and the
reinforcement design of beams varies among frames and stories according to the demands
calculated. Shear wall thickness is 0.30 meters and each leg of C-core fills an entire span, i.e.
each leg is 8 meters long. Two horizontal legs of the C-core lie on the x-axis (perpendicular to
direction of analysis) and they mainly provide lateral stiffness in that direction. Story height is
4 meters for the first story while it is 3.5 meters for the upper stories. Similar to the previous
case study, there is no basement and the structure is on fixed supports. The slabs have a
uniform thickness of 14 cm. Slab cover is considered as additional dead load with 1.5 kN/m2
intensity. Live load of 2 kN/m2 is also included in the design according to TS-498.
OpenSees software (2011) is employed in the modeling and analysis. In the modeling phase
of the structure, the main challenge exclusive to this case study is the analytical representa-
tion of shear walls. These members are modeled by using frame elements along the shear wall
axis, connected to the joints along the corner axes by rigid beams. The primary advantage of
this approach is its simplicity since it requires less computational effort, both in solution and
post processing. Another alternative that could be considered was employing shell elements
for shear walls. However, this modeling approach has many difficulties that need to be taken
into account. Firstly, shell elements are not used in practice as common as frame elements in
the OpenSees platform and its documentation is insufficient. As a result, verification of shell
models in 3D analysis environment is more difficult than frame element models. Also, com-
putation requirement in terms of effort and runtime accompanying shell elements is another
disadvantage that has to be considered. In the light of these issues, frame members with rigid
beam links are preferred in the modeling of shear walls. The model topology for shear walls
is shown in Figure 5.3.
After deciding on the shear wall model, linear elastic and nonlinear models have been pre-
pared for two different types of analyses. “ElasticBeamColumn” elements are employed for
beams, columns and shear walls in the linear elastic model.
In the nonlinear model, “beamwithHinges” elements are utilized for the beams and columns.
Moment-curvature relationships are defined for beams. Fiber sections with concrete and re-
inforcement material properties are defined for columns. Both of these sectional properties
75
Figure 5.3: Shear wall modeling approach employed in the case study.
are then assigned to the hinge length regions of “beamwithHinges” elements. The detailed
formulation of these sections and elements was discussed in the previous chapter. In the case
of shear walls, “nonlinearBeamColumn” elements are used rather than the “beamwithHinges”
element. Due to the nature of the deformation experienced by shear walls during seismic re-
sponse, plastic hinging is not confined to a certain length at the member ends; therefore an
element formulation with distributed plasticity model is required for accurate response calcu-
lation. As a distributed plasticity element, “nonlinearBeamColumn” members are available
in the OpenSees platform and consequently they are utilized in the modeling. By definition,
“nonlinearBeamColumn” elements require to specify a number of integration points along the
member where the response of the member itself is calculated. Five integration points along
these members are selected.
Shear walls in two directions have different reinforcement designs according to the demands
calculated. Also, along the critical shear wall height of Hcr = 14.5 meters (first four stories),
these designs have been modified according to the regulations stated in TEC. As a result, two
different sections have been defined for shear walls in both directions. The sections are mod-
eled as fiber sections with confined and core concrete properties and steel reinforcements as
in the case of columns, and assigned to the “nonlinearBeamColumn” elements that represent
76
shear walls.
Cracked sections for elastic frame elements are considered similar to what has been done in
the previous case study. In the linear elastic model, gross moments of inertia of beams are
multiplied with 0.4, that of columns are multiplied by 0.6. Gross moments of inertia of shear
walls are multiplied by 0.8. In the nonlinear model, cracked sections are introduced for the
elastic portions of “beamwithHinges” elements, with coefficients of 0.4 for beams and 0.6 for
columns. However, since distributed plasticity members with no elastic portions are used for
the shear walls, no cracked sections are defined for shear walls in the nonlinear model since
fiber sections automatically account for the gravity effects and stiffness of the members are
calculated accordingly.
P-Delta effects are included not only for columns, but also for the shear-walls. 5% damping
is estimated in the analysis where damping coefficients are obtained from the 1’st and 3’rd
modes.
Automation procedures are written in MATLAB (R2010b) where response history, response
spectrum, modal pushover and generalized pushover analyses are conducted for the ground
motion set given in Chapter 3. The general workflow in these written procedures closely fol-
lows the algorithm that is written for the analysis of the 8-Story building discussed in Chapter
4. Pre and post-process scripts are written in MATLAB as well as the codes that are generated
to run the OpenSees within MATLAB environment. Different from the work done in the pre-
vious chapter, 20-story building requires very complex procedures for pre and post-processing
and gathering of results since the number of frame members is large and a considerable large
amount of data is generated in the analyses. As a natural consequence of the increased model
size, runtime for analyses has also increased. Still, an average personal computer is capable
of performing these analyses which emphasizes the importance of adopting simplicity while
constructing the analysis model.
5.2 Free Vibration Properties
Modal properties of the building are presented in Table 5.1. Eigenvalue analysis is performed
for all sixty modes and the nine modes are presented in the table. The first column in Table 5.1
shows the mode number (X and Y stand for translation dominant mode whereas θ stands for
77
rotation dominant mode of the Y-θ couple). Corresponding period of each mode is displayed
in the second column. Effective modal mass (in tons) and the effective modal mass ratio for
X and Y directions are also presented.
Table 5.1: Free vibration properties of the twenty story structure
ModePeriod
(seconds)
EffectiveModal Mass
inY-direction
(M∗n,y) (tons)
EffectiveModal Mass
Ratio inY-direction
EffectiveModal Mass
inY-direction
(M∗n,x) (tons)
EffectiveModal Mass
Ratio inY-direction
1Y 2.36 6725.5 0.559 0 01X 2.10 0 0 7679.1 0.6381θ 1.06 813.98 0.068 0 02Y 0.64 1759.5 0.146 0 02X 0.38 0 0 2283.3 0.1902θ 0.36 133.54 0.011 0 03Y 0.23 870.8 0.072 0 03θ 0.20 316.8 0.026 0 03X 0.15 0 0 153.6 0.013
5.3 Issues Encountered During Analyses
Although simple material models and element formulations have been preferred over more
complex modeling options, the resulting analysis model is still rather complicated compared
with the 8-story structure. Different from the previous case study, presence of shear walls and
their connecting rigid beams in both directions has been one of the main source of modeling
challenge and complexity.
Convergence issues have been encountered in the analysis stage of this case study. Espe-
cially in the nonlinear dynamic time history analysis, the solution failed to converge under
most of the ground motions that have been employed. The poor performance of the model is
considered to depend partly on the inadequate variety of element models and documentation
in the OpenSees platform. For the linear elastic model, this was not the case since “elas-
78
ticBeamColumn” elements represented the behavior accurately enough. However, imposing
the actual shear wall behavior in a nonlinear model with the tools available in the OpenSees
platform had been a real challenge, and consequently the resulting analysis model suffered
from inadequacy in calculating the member forces.
Among the ground motion set discussed in Chapter 3, convergence has been achieved only
for half of the ground motion records. Results for these ground motions are presented in
the following section. Even though the global response of a mixed system (moment frames
and shear wall) is observed in the results (console deformation behavior becomes prominent
as frames gets closer to shear wall), the ground motions selected stressed the structure in a
limited sense.
5.4 Presentation of the Analysis Results
As stated in the previous section, convergence has been achieved under four ground motion
records, GM1, GM2, GM4 and GM5 given in Chapter 3. Results obtained from nonlinear
response history analysis (NRHA), generalized pushover analysis (GPA), modal pushover
analysis (MPA) and response spectrum analysis (RSA) of the twenty story structure are pre-
sented in Figures 5.4 to 5.19. Interstory drift ratios, mean beam plastic rotations, and mean
column chord rotations are plotted for the five frames in the structure in the direction of analy-
sis (denoted as FE, FI, CM, SI and SE). The chord rotations for the shear wall in the direction
of analysis are also plotted.
It is observed from the results that inelastic structural response to the ground motions is very
limited. Matching of GPA with NRHA is very good in GM2 results; it essentially captures the
dynamic response except for SI frame where shear wall is present. In the case of other ground
motion results, close performance of MPA and GPA is notable. This similar performance
could be due to fact that the first mode is dominant; therefore MPA is able to estimate the
fairly straightforward response by simply using the roof displacement. Since no or negligible
higher mode effects are apparent, two methods produce more or less the same results. When
low deformations are encountered, separation between the two may become less observable.
Overall, GPA successfully matches the mean beam plastic chord rotations especially in the
upper stories. Although some overshooting for lower stories is observed at the FE and FI
79
frames under GM5, the errors are minimal. It is also noticeable that the mean column chord
rotation results for all ground motions are closely matched for each analysis methods. GPA
yields the benchmark NRHA values accurately enough for column chord rotations.
Shear wall chord rotations are too small to make a reasonable comparison but GPA seems to
be successful in estimating the deformation trend observed in NRHA.
80
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
0 0.002 0.004 0.006 0.008 0.01
Story
#
Frame FE, GM1
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
0 0.002 0.004 0.006 0.008 0.01
Frame FI, GM1
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
0 0.002 0.004 0.006 0.008 0.01
Story
#
Interstory Drift Ratio
Frame CM, GM1
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
0 0.002 0.004 0.006 0.008 0.01
Interstory Drift Ratio
Frame SI, GM1
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
0 0.002 0.004 0.006 0.008 0.01
Story
#
Interstory Drift Ratio
Frame SE, GM1
NRHA GPA MPA RSA
Figure 5.4: Interstory drift ratios of five frames under GM1
81
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
0 0.002 0.004 0.006
Story
#
Frame FE, GM1
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
0 0.002 0.004 0.006
Frame FI, GM1
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
0 0.002 0.004 0.006
Story
#
Plastic Rotation (rad)
Frame CM, GM1
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
0 0.002 0.004 0.006
Plastic Rotation (rad)
Frame SI, GM1
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
0 0.002 0.004 0.006
Story
#
Plastic Rotation (rad)
Frame SE, GM1
NRHA GPA MPA
Figure 5.5: Mean beam plastic rotations of five frames under GM1
82
I Ends! J Ends!
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
0 0.001 0.002 0.003 0.004
Story
!#
Frame!FE,!GM1
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
0 0.0005 0.001 0.0015 0.002
Frame!FE,!GM1
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
0 0.001 0.002 0.003 0.004
Story
!#
Frame!FI,!GM1
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
0 0.0005 0.001 0.0015 0.002
Frame!FI,!GM1
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
0 0.001 0.002 0.003 0.004
Story
!#
Chord!Rotation!(rad)
Frame!CM,!GM1
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
0 0.0005 0.001 0.0015 0.002
Chord!Rotation!(rad)
Frame!CM,!GM1
NRHA GPA MPA
Figure 5.6: Mean column chord rotations of five frames under GM1
83
I Ends! J Ends!
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
0 0.001 0.002 0.003 0.004
Story
!#
Frame!SI,!GM1
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
0 0.0005 0.001 0.0015 0.002
Frame!SI,!GM1
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
0 0.001 0.002 0.003 0.004
Story
!#
Chord!Rotation!(rad)
Frame!SE,!GM1
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
0 0.0005 0.001 0.0015 0.002
Chord!Rotation!(rad)
Frame!SE,!GM1
NRHA GPA MPA
Figure 5.6: Mean column chord rotations of five frames under GM1 (continued)
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
0 0.00025 0.0005 0.00075 0.001
Story
#
Chord Rotation (rad)
Shear Wall I!End Chord Rotations, GM1
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
0 0.00025 0.0005 0.00075 0.001
Chord Rotation (rad)
Shear Wall J!End Chord Rotations, GM1
NRHA GPA MPA
Figure 5.7: Y-axis shear wall chord rotations under GM1
84
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
0 0.002 0.004 0.006 0.008 0.01 0.012
Sto
ry#
Frame FE, GM2
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
0 0.002 0.004 0.006 0.008 0.01 0.012
Frame FI, GM2
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
0 0.002 0.004 0.006 0.008 0.01 0.012
Sto
ry#
Interstory Drift Ratio
Frame CM, GM2
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
0 0.002 0.004 0.006 0.008 0.01 0.012
Interstory Drift Ratio
Frame SI, GM2
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
0 0.002 0.004 0.006 0.008 0.01 0.012
Sto
ry#
Interstory Drift Ratio
Frame SE, GM2
NRHA GPA MPA RSA
Figure 5.8: Interstory drift ratios of five frames under GM2
85
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
0 0.002 0.004 0.006
Sto
ry#
Frame FE, GM2
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
0 0.002 0.004 0.006
Frame FI, GM2
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
0 0.002 0.004 0.006
Sto
ry#
Plastic Rotation (rad)
Frame CM, GM2
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
0 0.002 0.004 0.006
Plastic Rotation (rad)
Frame SI, GM2
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
0 0.002 0.004 0.006
Sto
ry#
Plastic Rotation (rad)
Frame SE, GM2
NRHA GPA MPA
Figure 5.9: Mean beam plastic rotations of five frames under GM2
86
I-Ends J-Ends
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
0 0.001 0.002 0.003 0.004 0.005
Sto
ry#
Frame FE, GM2
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
0 0.0005 0.001 0.0015 0.002 0.0025
Frame FE, GM2
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
0 0.001 0.002 0.003 0.004 0.005
Sto
ry#
Frame FI, GM2
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
0 0.0005 0.001 0.0015 0.002 0.0025
Frame FI, GM2
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
0 0.001 0.002 0.003 0.004 0.005
Sto
ry#
Chord Rotation (rad)
Frame CM, GM2
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
0 0.0005 0.001 0.0015 0.002 0.0025
Chord Rotation (rad)
Frame CM, GM2
NRHA GPA MPA
Figure 5.10: Mean column chord rotations of five frames under GM2
87
I-Ends J-Ends
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
0 0.001 0.002 0.003 0.004 0.005
Sto
ry#
Frame SI, GM2
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
0 0.0005 0.001 0.0015 0.002 0.0025 0.003
Frame SI, GM2
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
0 0.001 0.002 0.003 0.004 0.005
Sto
ry#
Chord Rotation (rad)
Frame SE, GM2
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
0 0.0005 0.001 0.0015 0.002 0.0025
Chord Rotation (rad)
Frame SE, GM2
NRHA GPA MPA
Figure 5.10: Mean column chord rotations of five frames under GM2 (continued)
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
0 0.0005 0.001 0.0015
Sto
ry#
Chord Rotation (rad)
Shear Wall I-End Chord Rotations, GM2
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
0 0.0005 0.001 0.0015
Chord Rotation (rad)
Shear Wall J-End Chord Rotations, GM2
NRHA GPA MPA
Figure 5.11: Y-axis shear wall chord rotations under GM2
88
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
0 0.002 0.004 0.006 0.008
Sto
ry#
Frame FE, GM4
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
0 0.002 0.004 0.006 0.008
Frame FI, GM4
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
0 0.002 0.004 0.006 0.008
Sto
ry#
Interstory Drift Ratio
Frame CM, GM4
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
0 0.002 0.004 0.006 0.008
Interstory Drift Ratio
Frame SI, GM4
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
0 0.002 0.004 0.006 0.008
Sto
ry#
Interstory Drift Ratio
Frame SE, GM4
NRHA GPA MPA RSA
Figure 5.12: Interstory drift ratios of five frames under GM4
89
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
0 0.001 0.002 0.003 0.004 0.005
Sto
ry#
Frame FE, GM4
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
0 0.001 0.002 0.003 0.004 0.005
Frame FI, GM4
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
0 0.001 0.002 0.003 0.004 0.005
Sto
ry#
Plastic Rotation (rad)
Frame CM, GM4
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
0 0.001 0.002 0.003 0.004 0.005
Plastic Rotation (rad)
Frame SI, GM4
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
0 0.001 0.002 0.003 0.004 0.005
Sto
ry#
Plastic Rotation (rad)
Frame SE, GM4
NRHA GPA MPA
Figure 5.13: Mean beam plastic rotations of five frames under GM4
90
I-Ends J-Ends
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
0 0.001 0.002 0.003 0.004 0.005
Sto
ry#
Frame FE, GM4
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
0 0.0005 0.001 0.0015 0.002
Frame FE, GM4
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
0 0.001 0.002 0.003 0.004 0.005
Sto
ry#
Frame FI, GM4
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
0 0.0005 0.001 0.0015 0.002
Frame FI, GM4
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
0 0.001 0.002 0.003 0.004 0.005
Sto
ry#
Chord Rotation (rad)
Frame CM, GM4
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
0 0.0005 0.001 0.0015 0.002
Chord Rotation (rad)
Frame CM, GM4
NRHA GPA MPA
Figure 5.14: Mean column chord rotations of five frames under GM4
91
I-Ends J-Ends
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
0 0.001 0.002 0.003 0.004 0.005
Sto
ry#
Frame SI, GM4
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
0 0.0005 0.001 0.0015 0.002
Frame SI, GM4
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
0 0.001 0.002 0.003 0.004 0.005
Sto
ry#
Chord Rotation (rad)
Frame SE, GM4
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
0 0.0005 0.001 0.0015 0.002
Chord Rotation (rad)
Frame SE, GM4
NRHA GPA MPA
Figure 5.14: Mean column chord rotations of five frames under GM4 (continued)
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
0 0.0005 0.001 0.0015
Sto
ry#
Chord Rotation (rad)
Shear Wall I-End Chord Rotations, GM4
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
0 0.0005 0.001 0.0015
Chord Rotation (rad)
Shear Wall J-End Chord Rotations, GM4
NRHA GPA MPA
Figure 5.15: Y-axis shear wall chord rotations under GM4
92
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
0 0.002 0.004 0.006 0.008 0.01 0.012
Story
#
Frame FE, GM5
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
0 0.002 0.004 0.006 0.008 0.01 0.012
Frame FI, GM5
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
0 0.002 0.004 0.006 0.008 0.01 0.012
Story
#
Interstory Drift Ratio
Frame CM, GM5
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
0 0.002 0.004 0.006 0.008 0.01 0.012
Interstory Drift Ratio
Frame SI, GM5
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
0 0.002 0.004 0.006 0.008 0.01 0.012
Story
#
Interstory Drift Ratio
Frame SE, GM5
NRHA GPA MPA RSA
Figure 5.16: Interstory drift ratios of five frames under GM5
93
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
0 0.002 0.004 0.006
Story
#
Frame FE, GM5
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
0 0.002 0.004 0.006
Frame FI, GM5
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
0 0.002 0.004 0.006
Story
#
Plastic Rotation (rad)
Frame CM, GM5
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
0 0.002 0.004 0.006
Plastic Rotation (rad)
Frame SI, GM5
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
0 0.002 0.004 0.006
Story
#
Plastic Rotation (rad)
Frame SE, GM5
NRHA GPA MPA
Figure 5.17: Mean beam plastic rotations of five frames under GM5
94
I Ends! J Ends!
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
0 0.001 0.002 0.003 0.004 0.005
Story
!#
Frame!FE,!GM5
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
0 0.0005 0.001 0.0015 0.002 0.0025
Frame!FE,!GM5
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
0 0.001 0.002 0.003 0.004 0.005
Story
!#
Frame!FI,!GM5
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
0 0.0005 0.001 0.0015 0.002 0.0025
Frame!FI,!GM5
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
0 0.001 0.002 0.003 0.004 0.005
Story
!#
Chord!Rotation!(rad)
Frame!CM,!GM5
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
0 0.0005 0.001 0.0015 0.002 0.0025
Chord!Rotation!(rad)
Frame!CM,!GM5
NRHA GPA MPA
Figure 5.18: Mean column chord rotations of five frames under GM5
95
I Ends! J Ends!
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
0 0.001 0.002 0.003 0.004 0.005
Story
!#
Frame!SI,!GM5
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
0 0.0005 0.001 0.0015 0.002 0.0025
Frame!SI,!GM5
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
0 0.001 0.002 0.003 0.004 0.005
Story
!#
Chord!Rotation!(rad)
Frame!SE,!GM5
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
0 0.0005 0.001 0.0015 0.002 0.0025
Chord!Rotation!(rad)
Frame!SE,!GM5
NRHA GPA MPA
Figure 5.18: Mean column chord rotations of five frames under GM5 (continued)
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
0 0.00025 0.0005 0.00075 0.001
Story
#
Chord Rotation (rad)
Shear Wall I!End Chord Rotations, GM5
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
0 0.00025 0.0005 0.00075 0.001
Chord Rotation (rad)
Shear Wall J!End Chord Rotations, GM5
NRHA GPA MPA
Figure 5.19: Y-axis shear wall chord rotations under GM5
96
CHAPTER 6
SUMMARY and CONCLUSIONS
6.1 Summary
Generalized pushover analysis is developed in this study as a nonlinear static analysis proce-
dure for estimating the inelastic response of torsionally coupled structures under earthquake
ground motions. The analysis procedure is built upon the generalized force vector concept
which is defined as the effective force vector acting on the system at the instant tmax when
a specific response parameter reaches its maximum value. The specific response parameter
is selected as the interstory drift of j’th story (∆ j). Although this generalized force vector is
dependent on tmax in a true dynamic response history analysis, it is approximately defined in
terms of modal spectral responses obtained via response spectrum analysis in this study by
using a modal scaling rule. The rule is applied for interstory drift contribution of all modes
to the j’th interstory drift (∆i, j), which yields the generalized force vector in terms of modal
spectral responses. They are applied to the structure in an incremental form until the target
response parameter is attained. Target response parameter, which is the interstory drift, is
also calculated for all stories by using the same modal scaling rule. By considering the nature
of the response of unsymmetrical plan structures, this formulation is further extended to a
frame-by-frame approach where different target drift demands and generalized force vectors
are determined for each frame at each story. It is observed that in an unsymmetrical-plan
building, different frames on the same story attain their tmax values at different time instants.
Therefore the aforementioned approach which accounts for this phenomenon is expected to
provide better response estimation. A practical implementation of the procedure is also dis-
cussed where the structure is pushed with generalized force vectors defined at the center of
mass meanwhile the interstory drift demands for the individual frames are employed as tar-
97
get demand parameters. This simplification requires less computational effort and runtime
while providing the results with the same accuracy. At the final stage of the analysis, results
gathered for each frame from the pushover analyses under all generalized force vectors are
combined in an envelope algorithm.
Two case studies are prepared in order to evaluate the performance of the proposed analysis
procedure. The first one is an eight story moment frame structure with mass eccentricity
while the second one is a twenty story frame-wall building where the shear wall core is placed
with offset from the center of mass. Both of the structures are analyzed under eight ground
motions. Nonlinear response history, generalized pushover, modal pushover and response
spectrum analyses are performed in both case studies. Interstory drift ratios, mean beam
plastic rotations and mean column chord rotations of each frame at each story obtained from
different analysis procedures are compared. Story shear forces, second and sixth story beam
end moments of each frames are also compared for the eight story structure.
6.2 Conclusions
According to the results obtained in this study, the following conclusions are reached:
• For the majority of the ground motion records, GPA is successful in estimating the
response of eight story structure compared to benchmark NRHA. Higher mode effects
are well captured and this is best observed in interstory drift ratios and mean beam
plastic rotations. It is also noted that if matching of RSA and LERHA (Linear Elastic
Response History Analysis) is good enough, then GPA yields very good estimates since
GPA includes RSA as a prerequisite. Using the inelastic spectral displacement demand
D∗n for the first mode couple results in an improvement for some of the ground motions.
• GPA is very accurate in estimating the internal forces. No modal combination is em-
ployed while compiling the results at the final step of GPA, therefore it yields the forces
directly as they were recorded during analysis. Both of the case studies are designed
with capacity design principles and this is observed at the beam end moment compar-
isons. NRHA and GPA give almost the same moment values for most of the cases
since capacity of the beam member is reached in both analyses. However, other anal-
ysis methods that employ modal combination rule overestimate the force values and
98
exceed member capacities, which is unrealistic. Story shear forces estimated by GPA
also match the NRHA results well, which underlines the success in force estimation.
• Modeling challenges that is encountered in the twenty story frame-wall structure made
the analysis difficult. For the four ground motions where convergence had been achieved,
member deformations were small. The massive shear wall core gives a high lateral stiff-
ness to the structure and the global response stayed at minimum. Also torsionally stiff
behavior of the structure did not result in prominent higher mode effects. Although the
recorded deformations were limited, GPA was successful in estimating the response
under ground motions where the higher mode effects were significant. However shear
wall rotations were very small and it was difficult to make an evaluation of the proposed
procedure based on these values. In general, it can be suggested that GPA has the poten-
tial to analyze complex high-rise structures which are located on the long period range
of the response spectrum and GPA is successful in capturing the global deformation
trend of the system. However, this needs to be evaluated with analysis models that are
prepared on a more sound software environment that has the necessary tools to model
these complex structures.
• Lack of convenient material models and element formulations for 3D modeling of shear
walls is an issue in the OpenSees platform. Due to this limitation, approximate model-
ing solutions that are employed may not reflect the actual behavior properly as discussed
above. Also issues with rigid diaphragms and fiber sections which result in unrealisti-
cally high axial loads in beams when fiber sections are used is another drawback of the
software. These problems should be further investigated and taken into consideration
while modeling with OpenSees.
• The potential of GPA as an approximate analysis method is noticeable when the analysis
results are investigated. It is a good alternative to the nonlinear response history analysis
and its estimations for unsymmetrical-plan structures are generally acceptable. For long
period systems, its accuracy may further be tested in different analysis environments
that are capable of representing the behavior of structural members like shear walls
successfully in a 3D model.
99
REFERENCES
F.S. Alıcı, K. Kaatsız, and H. Sucuoglu. "Genel Yük Vektörleri Ile Çok Modlu Itme Anal-
izi (Genel Itme Analizi)". Birinci Türkiye Deprem Mühendisligi ve Sismoloji Konferansı,
Ankara, Turkey, 2011.
S. Antoniou and R. Pinho. "Advantages and Limitations of Adaptive and Non-Adaptive
Force-Based Pushover Procedures". Journal of Earthquake Engineering, Volume 8 (No.
4): pp. 497–522, 2004a.
S. Antoniou and R. Pinho. "Development and Verification of a Displacement Based Pushover
Procedure". Journal of Earthquake Engineering, Volume 8 (No. 5): pp. 643–661, 2004b.
Applied Technology Council (ATC). "Seismic Evaluation and Retrofit of Concrete Build-
ings", Report No. ATC-40, Volume 1-2, 1996.
R.C. Barros and R. Almeida. "Pushover Analysis of Asymmetric Three-Dimensional Building
Frames". Journal of Civil Engineering and Management, Volume 11 (No. 1): pp. 3–12,
2011.
J.M. Bracci, S.K. Kunnath, and A.M. Reinhorn. "Seismic Performance and Retrofit Evalua-
tion of Reinforced Concrete Structures". Journal of Structural Engineering, ASCE, Volume
123 (No. 1): pp. 3–10, 1997.
A.M Chandler and X.N. Duan. "Performance of Asymmetric Code-Designed Buildings for
Serviceability and Ultimate Limit States". Earthquake Engineering and Structural Dynam-
ics, Volume 26 (No. 7): pp. 717–735, 1997.
C. Chintanapakdee and A.K. Chopra. "Evaluation of Modal Pushover Analysis Using Generic
Frames". Earthquake Engineering and Structural Dynamics, Volume 32: pp. 417–442,
2003.
C. Chintanapakdee and A.K. Chopra. "Seismic Response of Vertically Irregular Frames: Re-
sponse History and Modal Pushover Analyses". Journal of Structural Engineering, ASCE,
Volume 130 (No. 8): pp. 1177–1185, 2004.
100
A.K. Chopra. "Dynamics of Structures: Theory and Applications to Earthquake Engineer-
ing", 2nd Edition. Prentice Hall, Englewood Cliffs, N.J., 2001.
A.K. Chopra and R.K. Goel. "A Modal Pushover Analysis Procedure for Estimating Seismic
Demands for Buildings". Earthquake Engineering and Structural Dynamics, Volume 31:
pp. 561–582, 2002.
A.K. Chopra and R.K. Goel. "A Modal Pushover Analysis Procedure to Estimate Seismic
Demands for Unsymmetric-plan Buildings". Earthquake Engineering and Structural Dy-
namics, Volume 33: pp. 903–927, 2004.
Comite European de Normalisation. "European Standard EN 1998-3:2005 Eurocode 8. De-
sign of Structures for Earthquake Resistance - Part 3: Assessment and Refrofitting of Build-
ings", 2005.
P. Fajfar. "A Nonlinear Analysis Method for Performance-Based Seismic Design". Earth-
quake Spectra, Volume 16 (No. 3): pp. 573–592, 2000.
P. Fajfar and M. Fischinger. "N2 - A Method for Nonlinear Seismic Analysis of Regular Build-
ings". Proceedings of 9th World Conference on Earthquake Engineering, Tokyo-Kyoto,
Japan, Volume V: pp. 111–116, 1988.
P Fajfar and P. Gaspersic. "The N2 Method for the Seismic Damage Analysis of RC Build-
ings". Earthquake Engineering and Structural Dynamics, Volume 25: pp. 31–46, 1996.
P. Fajfar, D. Marusic, and I. Perus. Torsional Effects in the Pushover-Based Seismic Analysis
of Buildings. Journal of Earthquake Engineering, Volume 9 (No. 6): pp. 851–854, 2005.
Federal Emergency Management Agency. "FEMA-356, Prestandard and Commentary for the
Seismic Rehabilitation of Buildings", prepared by American Society of Civil Engineers for
the Federal Emergency Management Agency, 2000.
S.A. Freeman, J.P. Nicoletti, and J.V. Tyrell. "Evaluation of Existing Buildings for Seismic
Risk - A Case Study of Puget Sound Naval Shipyard, Bremerton Washington". Proceedings
of the First U.S. National Conference on Earthquake Engineering, Berkeley, California,
1975.
R.K. Goel and A.K. Chopra. "Extension of Modal Pushover Analysis to Compute Member
Forces". Earthquake Spectra, Volume 21 (No. 1): pp. 125–139, 2005.
101
B. Gupta and S.K. Kunnath. "Adaptive Spectra-Based Pushover Procedure for Seismic Eval-
uation of Structures". Earthquake Spectra, Volume 16 (No. 2): pp. 367–391, 2000.
S. Jerez and A. Mebarki. "Seismic Assessment of Framed Buildings: A Pseudo-Adaptive
Uncoupled Modal Response Analysis". Journal of Earthquake Engineering, Volume 15
(No. 7): pp. 1015–1035, 2011.
E. Kalkan and S.K. Kunnath. "Adaptive Modal Combination Procedure for Nonlinear Static
Analysis of Building Structures". Journal of Structural Engineering, ASCE, Volume 132
(No. 11): pp. 1721–1731, 2006.
D.C. Kent and R. Park. "Flexural Members with Confined Concrete". Journal of the Structural
Division, Volume 97 (No. 7): pp. 1969–1990, 1971.
V. Kilar and P. Fajfar. "Simple Push-Over Analysis of Asymmetric Buildings". Earthquake
Engineering and Structural Dynamics, Volume 26 (No. 2): pp. 233–249, 1997.
A. Kosmopoulos and M.N. Fardis. "Seismic Evaluation of Strongly Irregular and Torsionally
Unbalanced Concrete Buildings". Proceedings of the 2nd FIB Congress, Naples, Italy,
2006.
H. Krawinkler and G.D.P.K. Seneviranta. "Pros and Cons of a Pushover Analysis of Seismic
Performance Evaluation". Engineering Structures, Volume 20: pp. 452–464, 1998.
M. Kreslin and P. Fajfar. The Extended N2 Method Taking into Account Higher Mode Effects
in Elevation. Earthquake Engineering and S, Volume 40 (No. 14): pp. 1571–1589, 2011.
M. Kreslin and P. Fajfar. "The Extended N2 Method Considering Higher Mode Effects in
Both Plan and Elevation". Bulletin of Earthquake Engineering, Volume 10: pp. 695–715,
2012.
J.A. Mahaney, T.F. Paret, B.E. Kehoe, and S.A. Freeman. "The Capacity Spectrum Method
for Evaluating Structural Response During the Loma Prieta Earthquake". Proceedings of
the 1993 National Earthquake Conference, Memphis, pages pp. 501–510, 1993.
J. Mao, C. Zhai, and L. Xie. "An Improved Modal Pushover Analysis Procedure for Estimat-
ing Seismic Demands of Structures". Earthquake Engineering and Engineering Vibration,
Volume 7 (No. 1): pp. 25–31, 2008.
102
Mathworks Inc. MATLAB - R2010b, 2010.
Ministiry of Public Works and Settlement. "Turkish Earthquake Code: Specifications for the
Buildings to be Constructed in Disaster Areas", 2007.
A.S. Moghadam and W.K. Tso. "Pushover Analysis for Asymmetrical Multistorey Build-
ings". Proceedings of the 6th U.S. National Conference on Earthquake Engineering, EERI,
Oakland, California, 1998.
A.M. Mwafy and A.S. Elnashnai. "Static Pushover versus Dynamic Collapse Analysis of RC
Buildings". Engineering Structures, Volume 23: pp. 407–424, 2001.
OpenSees. "Open System for Earthquake Engineering Simulation".
htttp://opensees.berkeley.edu, November 2011.
T.F. Paret, K.K. Sasaki, D.H. Eilbeck, and S.A. Freeman. "Approximate Inelastic Procedures
to Identify Failure Mechanisms From Higher Mode Effects". 11th World Conference on
Earthquake Engineering, Paper No. 966, Acapulco, Mexico, 1996.
A. Reinhorn, A. Rutenberg, and J. Glück. "Dynamic Torsional Coupling in Asymmetric
Building Structures". Building and Environment, Volume 12 (No. 4): pp. 251–261, 1977.
J.C. Reyes and A.K. Chopra. "Three-Dimensional Modal Pushover Analysis of Buildings
Subjected to Two Components of Ground Motion, Including Its Evaluation for Tall Build-
ings". Earthquake Engineering and Structural Dynamics, Volume 40 (No. 7): pp. 789–806,
2011.
F.R. Rofooeia, N.K. Attaria, A. Rasekhb, and A.H. Shodjaa. "Adaptive Pushover Analysis".
Asian Journal of Civil Engineering (Building and Housing), Volume 8 (No. 3): pp. 343–
358, 2007.
A. Rutenberg. "Nonlinear Response of Asymmetric Building Structures and Seismic Codes:
A State of the Art Review". Nonlinear Seismic Analysis and Design of Reinforced Concrete
Buildings, pages pp. 281–305, 1992.
A. Rutenberg. "EAEE Task Group (TG) 8: Behaviour of Irregular and Complex Structures:
State of the Art Report: Seismic Nonlinear Response of Code-Designed Asymmetric Struc-
tures". Proceedings of the 11th European Conference on Earthquake Engineering, Paris,
France, 1998.
103
A. Rutenberg and A.C. Heidebrecht. "Approximate Analysis of Asymmetric Wall-frame
Structures". Building Science, Volume 10 (No. 1): pp. 27–35, 1975.
A. Rutenberg, A.M. Chandler, X.N. Duan, and J.C Correnza. "Nonlinear Seismic Response
of Asymmetric Structures: Bibliography". National Building Research Institute, Haifa,
1995.
M. Saiidi and M.A. Sözen. "Simple Nonlinear Seismic Analysis of R/C Structures". Journal
of the Structural Division, ASCE, Volume 107 (No. ST5): pp. 937–952, 1981.
K.K. Sasaki, S.A. Freeman, and T.F. Paret. "Multimode Pushover Procedure (MMP)-A
Method to Identify the Effects of Higher Modes in a Pushover Analysis". Proceedings
of the 6th US National Conference on Earthquake Engineering, Seattle, Washington, 1998.
K.G. Stathopoulos and S.A. Anagnostopoulos. "Inelastic Earthquake Response of Single-
Story Asymmetric Buildings: An Assessment of Simplified Shear-Beam Models". Earth-
quake Engineering and Structural Dynamics, Volume 32 (No. 12): pp. 1813–1831, 2003.
H. Sucuoglu and M.S. Günay. "Generalized Force Vectors for Multi-mode Pushover Anal-
ysis". Earthquake Engineering and Structural Dynamics, Volume 40 (No. 1): pp. 55–74,
2011.
Turkish Standard Institute. "Design Loads for Buildings", November 1997.
Turkish Standard Institute. "Requirements for Design and Construction of Reinforced Con-
crete Structures", February 2000.
S. Wilkinson and D. Thambiratnam. "Simplified Procedure for Seismic Analysis of Asym-
metric Buildings". Computers and Structures, Volume 79 (No. 32): pp. 2833–2845, 2001.
E. L. Wilson, A. Der Kiureghan, and E. P. Bayo. "A Replacement for the SRSS Method in
Seismic Analysis". Earthquake Engineering and Structural Dynamics, Volume 9 (No. 2):
pp. 187–192, 1981.
P Yang and Y. Wang. "A Study on Improvement of Pushover Analysis". Proceedings of 12th
World Conference on Earthquake Engineering, New Zealand, No. 1940: pp. 1–8, 2000.
104